Network failure technology Some Dongdong

For components that general users do not use, such as accessories, code conversion, etc. can be deleted. If you want to delete these unwanted components in Windows 98, you can open the Control Panel, Click to Add / Remove Programs. "Icon, select and delete in the Add / Delete Windows Components tab. However, in Windows 2000, it is not possible to delete this method because there is no such components in the Add / Remove Windows Component list. Is there a way to add these components in the Add / Remove Programs list? ---- Introduction to a method that can add these components in the Add / Remove Windows Components list, then remove these unnecessary components, the specific steps are as follows. ----1. Double-click the "My Computer" icon on your desktop, select Tools * "Folder Selection", select "Show all files and folders" in the Advanced Settings option of the View tab. The purpose of this setting is to display hidden folders and files. ----2. Go to the Windows 2000 system directory, generally for the winnt directory, then enter the inflicity (that is, C: / WinNT / INF), this directory is an implicit directory, there is a sysoc.inf file in this directory, this file is only 4KB, first Back up this file, use the text editing software to open this file after the backup is complete, then click the "Edit" menu, select the "Replace" option, enter ", hide" in "Finding Content", what is not all in "replacement" Lose, then click the "All Replace" buttons to save this file exit. ---- 3. Click Start * "Settings" * "Control Panel", double-click the Add / Remove Programs icon, then click the Add / Remove Windows Component button, open the Windows Component Wizard window, more than 5 in the component Items, such as "COM ", "Distributed Transaction Coordinator", "Fax Services", "Accessories, Tools", "Image Processing", etc. (as shown in the drawings), then remove the components from it is deleted. -------------------------------------------------- ----- 17, RIS client pre-classification and run sysprep.exe You can use pre-prix when using RIS to install Windows 2000 Professional, you must use the pre-prix to enter the client's GUID, how to enter? A global unique identifier (GUID) that relies on the computer. GUID is usually the MAC address of the NIC. In front of the MAC address of the NIC, add enough forward zero to ensure that the GUID is 32 16-based numbers is a complete GUID. Example 000000-0000-0000-0000-5254AB15C2EE wherein "52-54-AB-15-C2-EE" is the MAC address of the client network card. After running sysprep.exe on an experimental machine, you will be prompted to turn off your computer. At this time, it should be noted when you restart your computer. If you still start from your hard drive, you will be installed. So you should choose from to start the computer from the floppy disk or other media, and use a third-party image software to make clones throughout the hard disk or partition. Then distribute the cloned files into each computer, then use the third-party image software to solve the image on each computer, then select the installation from the hard disk.

Before running sysprep, let's first familiarize with its main switch parameters: -sysprep.inf Use sysprep.inf to configure a computer, apply to a computer-nosidgen computer that is used to clone the same configuration, does not generate a new SID number -PNP computer after reboot Reproducing PNP Device-Reboot runs sysprep restarts your computer, not shutdown ------------------------------------------------------------------------------------------------------------------------------------------------------------ --------------------- 18, using RIS to implement the client to install the operating system: If you start with the NIC, first ensure the client's system BIOS has put the network card Set to the first startup device. When the client starts, it automatically connects the RIS server. After the connection is connected, the RIS server prompts the user to press the F12 to download the user to install the Smart Wizard, and the user only needs to be repaired when the first installation system or system is damaged. F12, It is usually not used. If you start with a floppy disk, you will be prompted to press the F12 to initialize the network service. After Press F12, you must remember to remove the startup floppy before the character appears. -------------------------------------------------- -19 How to upgrade NT4 domain to Win2000 domain, you must pay attention to the following: 1, NT4's primary domain controller PDC must be used in NTFS file system format, if the FAT16 format will fail; 2, NT4 upgrade is unidirectional, So there is no way to upgrade Win2000B3 will return to the NT4 environment; 3. When doing upgrade planning, you must first back up important information and pre-preparation work, and remove applications that cannot be performed in Win2000B3, etc.; 4, First upgrade the PDC and upgrade BDC (before upgrading PDC, first upgrade the BDC offline, wait until the Win2000B3 domain controller, then upgrade the BDC or select the environment that does not upgrade the BDC to retain NT4); 5, upgrade, installation There are three important directories that need to be specified, including Active Directory, log profiles, and sysvol directories, determine the path of the directory under the NTFS partition; 6, the user account, group account and computer account will remain. -------------------------------------------------- --20, in Win2000, the fax first, first install a modem that supports faxes normally; after installing the modem, the modem you use will appear in the lower right corner of the taskbar. Since the Fax machine of Win 2000 is sent without receiving fax, if we want it to automatically receive the fax, you must right-click your modem name and select in the pop-up menu. "Receive"; additional computer transmission fax is implemented by "print" function, whether it is the same regardless of the image or text, so we can pass the content of the fax in Word or other editor, then select File / Print "Menu, select" Fax "and click" Print "button," Fax Send Wizard ", click" Next ", enter the recipient's name and phone number, and then select" Immediately according to its own actual situation "or "Time to send". ------------------------------------------- ------------ 21, first install Windows XP and then install Windows 2000 After installing the startup menu to enter the restore consumption, run the command: fixmbr C: fixboot C: Copy x: / i386 / ntldr C: COPY x: /i386/ntdetect.com C: X is the optical drive.

-------------------------------------------------- - - 22 Error information: "The Windows Installer Service Could Not Be Accur if you are running Windows in Safe Mode, or if The Windows Installer IS Not Correctly Installed ....". The reason for this error is Windows 2000 SP3 Contains version 2.0.2600.2 version of Windows Installer, when DCOM's default IMPERSOTION Level is set to Anonymous, this MSI version cannot run normally. And after the W2K SP3 is installed, the system will remain in the system, even if you re-upgrade SP3, it will not function properly. In order to successfully install SP3, you need to set the Impersonation Level for the DCOM for Identify, and to delete the residual file /%Windir%/system32/msisip.dll when the installation fails. DCOM Impersonation Level Setting: 1. Run DCOMCNFG 2. If you have some objects are not registered, the program will prompt you to register. 3. Enter the DCOM Setting Properties, select Default Properties, and set the default impersonation for Identify. -------------------------------------------------- ----- 23, making custom boot floppy disk 1, use Format A: / s in pure DOS to process a floppy disk 2, create file config.sys in the floppy disk, and edit the command line: device = oakcdrom.sys / d: mscdxxx device = himem.sys (for system information file Himem.sys), so that the DOS system can run smartdrv.exe) DOS = high, umb files = 60buffers = 803, create file autoexec in the disk. BAT, and edit the command line: LoadHigh A: / smartdrv / l A: /mscdex.exe /d: mscdxx4 ,oakcdrom.sys and himem.sys Copy to a disk Note: Command line in config.sys Device = OakCDrom.sys / d: Mscdxxx and autoexec.bat's command line: A: /Mscdex.exe / d: MSCDXXX is to load the optical drive under pure DOS. The xxx in the MSCDXXX in config.sys and autoexec.bat can be any character, but both must be the same. -------------------------------------------------- --- 24, handmade a route in the system and make it permanently curing, for example, "Route Add 192.168.50 192.168.255.255.0 192.168.4.100, but how to automatically load the route each time, can be added back, then add a parameter -P It became a permanent curing.

After adding a static routing record using the -p parameter, this record is stored in the registry, in the specific location: hkey_local_machine / system / currentcontrolset / services / tcpip / parameters / persistentRoutes -------------- ----------------------------------------- 25, Win2000 for the for command application: Open a DOS window in Win2000, then perform FOR / L% a in (1,1,254) do start / min / low telnet 192.168.0.% A 3389, all open 3389 port hosts After exposing this command, you will open 254 small windows in the taskbar. It is recommended to have no archive, save it. Scan a host's multiple ports, as follows for / l% a in (1, 1,65535) Do START / LOW / MIN Telnet 192.168.0.1% A This scans 192.168.0.1 1 to 65535 ports to scan all ports for / l% a in (1, 1, 254) do for / l% B IN (1,1,65535) do start / low / min telnet 192.168.0.% b This will scan all 1 to 65535 segments of 192.168.0.x paragraphs or above can only be used under Win2000, Because / l accumulation parameters are Win2000 to the extension of For course WinXP and Win.Net can be used. Other have not tried it yet ---------------------- ----------------------------- 26, MTU We know that the MTU (Maximum Transfer Unit Maximum Transmission Unit) defines TCP / IP protocols The number of bytes included in the network transmitted in the network can effectively improve the access speed and stability of the network by adjusting this parameter. In a Windows95 environment, the default setting of this value is 1500 bytes. Since the MTU is set to 576 bytes of routers, the result of the data packet size has caused 1500 bytes of data packets. Slimed into a number of 576 bytes or less packets, causing the system transfer speed to decrease. In order to match the size of the packet, you must regulate the relevant numerical value under the registration. -------------------------------------------------- --27, Windows serial number in the operating system WINDOWS 98 installation serial number: 1, open the registry editor under the premise of enter the operating system, expand HKEY_LOCAL_MACHINE / SOFTWARE / Microsoft / Windows / CurrentVersion branch, right Find the "ProductKey" string value in the side window, and its key value is the serial number! 2, if you can't enter the operating system, you can view the serial number of Windows 98 under DOS: Start your computer with the boot disk, then enter C: / Windows / Command / Find / L "ProductKey" C: /Windows/system.dat (assuming C The disk is a system disk), and the serial number of Windows 98 will be displayed after entering. WINDOWS 2000 installation serial number: Enter the Windows 2000 system installation folder, open the / system32 folder, find $ winnt $ .inf file, double-click the notepad to open, [userData] section "ProductID =" back quotation characters String is the serial number of Windows 2000! XP and 2003 are the same.

---------------------------------------------- 28, SYS .Comsys.com command, this command is specifically used to "transfer" from the source disk to the target disk, and will copy the system file on the source disk (such as IO.SYS and other system files) to the corresponding position of the target disk. Make the target disk can be started separately. When you turn it on, the system will prompt you: Invalid system disk, MS-DOS version does not match the error message, the command.com command is damaged, you can "pass the correct operating system" pass "to your hard disk, such as It turns out that your computer is started from the C drive, then you can type "SYS C:" after starting with the boot disk. After the operation is completed, the system will prompt: "System Transferred", after restarting, it is generally possible to run normally, so it is removed from your reinstall system. The SYS command is the cleansing means of rebuilding the Boot area, or killing the Boot area virus. (The Boot area here is actually the guiding sector on the activity partition) --------------------------------- --------------- 29, FDISK hidden parameters switch its grammatical format: fdisk / parameter 1, / ACTOK does not check if there is a bad sector in disk surface: Use this parameter to speed up partition operating. 2, / cMBR Re-establish the main boot record of the specified disk: its operation is equivalent to / MBR parameters, and the difference is that the disk can be specified. 3, / EXT Establish extended partition: establish an extension partition on the current disk (establish a logical disk partition for aged) 4, / fprmt In interactive mode U. Support for large-capacity inquiry screens, but inquiry when establishing a new partition, use FAT16 or FAT32 format. 5, / log build disk logical partition: establish a logical disk, must be used at the same time as / ext. 6, / logo uses FAT16 to establish a disk logical partition. 7. / MBR Re-establishs the main boot record 446 bytes of the master boot record. 8, / PRI establishes a residential area, and the partition will automatically be set to an activation form. 9. / Prio uses the FAT16 format to establish a primary partition and set to an activation form. 10, / Q End FDisk is not restarting the computer: Use it to restart your computer after changing the partition table. 11, / status Display Current Disk Partition Status: When there is no logical partition on the disk extension partition, the extension partition will not be displayed. 12, / x Do not use the LBA property: Use it will not produce partitions with LBA properties. -------------------------------------------------- -30, completely delete WinMe's system restore 1. First click on "My Computer" to select "Properties", in the Performance tab, select File System, "Disable System Restore" in the Troubleshooting tab Optional. 2, run the regedit, find hkey_local_machine / system / currentControlSet / Services / VXD / VXDMON / Clear all the keys 3, restart your computer to delete the C: / _ restore directory (hidden folder).

----------------------------------------- 31, Windows 2000 BUGS in a gives Everyone At the end of the directory of Full Control, you will create a file, remove the permission of this file, set to Administrator Full Control Only. Then use any ordinary user to log in, enter this EVERYONE FULL Control directory, try to access this file, of course, is of course refusing access. But you can delete it! This problem briefly explains that you can see through the description of WinHelp's File Permission. It is more referred to, and there is a relationship with a File Delete Child's hidden authority. Microsoft's explanation is to ensure that UNIX is compatible. Because the Win2K system is too large, if the permissions between the domains are more complicated, it will be very complicated. Therefore, if there is something that needs to be protected, perhaps use DENY permissions better. EFS is indeed a very powerful set of encryption systems, because of its power, we have to be cautious and cautious in use. First before the addition and deletion of AD, you must remove all EFS encryption, otherwise this EFS encrypted file or folder will never be opened. This is actually warned when installing or deleting the AD, can be easily ignored by the user. For specific reasons I think is because Local Administrator and the SID of Domain Administrator are different, so there is a problem. In addition, when performing Duplicate operation, it is necessary to remove EFS before executing sysprep, otherwise these data will never be read. Another problem with EFS is that it cannot prevent a file from being deleted. That is, as long as a user has permission to enter the directory where the EFS encryption file is located, you can delete it, so you can put it on the folder encryption and setting permissions in the EFS file. ------------------------------------------ 32, find the bottleneck 1 of the system, Find a memory bottleneck to use the following counter to identify a bottleneck memory resource: (1) System / Processor Queue Length (2) Memory / Pages / Sec 2, Find Disk Bottlenecks in Performance Tools Use the following counter to identify a bottleneck disk Resources: (1) PhysicalDisk /% Disk Time and% IDLE TIME (2) PhysicalDisk / Disk Reads / Sec and Disk Writes / Sec (3) PhysicalDisk / Avg.disk Queue Length (4) LogicalDisk /% Free Space also monitoring memory Counters to determine if there are too many memory pagings to make the disk used tensions. Note: Unlike physical disk counters, the data of logical disk counters is not collected by the operating system by default. To obtain a logical drive or the performance counter data of the storage volume, you must type diskperf -yv at the command prompt. This results in a disk performance statistical driver for collecting disk performance data to report data of logical drives and storage volumes. By default, the operating system uses the diskperf -yd command to include physical drive data. For more information on using the DiskPerf command, type Diskperf at the command prompt -?

3. Finding processor bottlenecks using the following counter to identify processor resources: (1) Processor / Interrupts / Sec (2) Processor /% Processor Time (3) Process /% Processor Time (4 System / Processor Queue Length 4, find network bottlenecks in performance tools to identify bottleneck network resources: (1) Network interface / bytes Total / Sec, Bytes Sent / Sec and Bytes Received / Sec (2) Protocol_Layer_Object / Segments Received / Sec, SEGMENTS SENT / Sec, Frames Sent / Sec, Frames Received / Sec, for the NWLINK performance object, with frame-related counters report zero. Use the object-based counter. (3) Server / Bytes Total / Sec, Bytes Received / Sec and Bytes Sent / Sec (4) Network Segment /% Network Utilization ---------------------- ---------------------- 33, there are many ways to close the 445 port, but I recommend the following methods: Modify the registry, add a key value Huest: hkey_local_machinekey: system / controlset / services / nett / parametersName: SMBDEVICEENABED TYPE: REG_DWORDVALUE: 0 After the modification, restart the machine, run "netstat -an", you will find that your 445 port is no longer listenging. -------------------------------------------- 34, Windows 98 Extract.exe and ext.exe actually, the Windows 98 super boot disk also has this command, this already appearing in DOS6.22, originally only the Microsoft's development of a command to decompress the CAB compression package, but it The function is more than just decompression, we are more for using it to restore system files in the CDD. (If you can enter the graphical interface, you can use SFC, which is more powerful, convenient) The following is the use parameter and command format of the extract.exe command: / y Overwrite the existing files no longer prompt / a process all CAB File, from the specified first CAB file start / D display the directory of the file contained in the CAB file, no decompression / e decompression, replace "*. *", That is, unpacking all file / L DIR specified solutions The location of the compressed file, default is the current directory / c copy CAB file? K When we know that the system files you want to recover can be used when you install the location, you can use the extract / y win98_27.cab cdfs.vxd command (from Windows 98's installation CD named Win98_27.cab compression package Recovered CDFS.VXD files). However, if you don't know the exact location at this file, you must first use the extract / d win98 _ *. CAB (here * is a number representing the CAB file package) to view CDFS.VxDG for each CAB file, Trouble. Ext.exe.

I just introduced that there is a bit of trouble when extract.exe finds a system file that the path is unstuten, if it combines it with ext.exe, you don't need to know the specific path to the file. However, it is important to note that this ext.exe command is in an ebd.cab compression package, you can't use it directly. You must first use the extract / y ebd.cab ext.exe command to extract EXT.EXE in the EBD.CAB compression package to the same directory, now you know the effect of EBD.CAB compression package? . ? K? K Use the ext.exe command to quickly restore the unknown path in the CD, type the ext.exe command, then "please enter the path Windows Cab Files" _ ", type Windows 98 Path where the disk, such as "G: / PWIN98", when the Enter "please enter to name (s) of the file (s) You want to extract:", type the name you need to find, if It is still cdfs.vxd and will prompt "Please enter the path fo extract", "Enter" for current Directory: "It means that you need to enter the path to be unzipped, such as C: / Windows / System ", will appear as follows after entering the bus again: The Options you entered WERE: Source Path: g: / pwin98file (s) to extract: cdfs.vxddestination path: c: / windows / systemis this okay? (Y / N): "Y", the ext command will start working quickly, but it can not only find the detailed location of the file, but can automatically decompress the file to the specified path, the speed is extremely fast. -------------------------------- ----------------------------- 35, is there a tool to drop the dynamic disk back to the basic disk? Try the following: Successful: Install Recovery Console, enter the recovery console, remove one or two unimportant volumes, restart with the diskpart command, discovery has changed from the dynamic disk back to the base plate, the data that is not deleted is intact ( Xiao Shuai did not have done !!!!!!) -------------------------------------- --- 36, Win2000 unconnected, IP unavailable solution Win2000 has such a feature, when our network cable is disconnected due to various reasons, you will prompt your network cable without connecting, you will ping your own The address is not powerful, causing some of the tests that only the net cards do not even connect to the nets are not can't do. Now give a solution, but not guaranteed certain success (this is unsuccessful on a large part of the NIC. Win2000 The function is called "Media Sense) 1. We can prohibit this feature through the registry: hkey_local_machine / s YSTEM / CURRENTCONTROLSET / SERVICES / TCPIP / Parameters There is a keyword for disabledhcpmediasense, DWORD type (if you don't have yourself), the default is 0 to open the detection function, set it to 1 to turn off the probe. After the modification is complete, restart the computer, then you remove the network cable, though the system is still not connected, but the IP bound to the NIC is still tube.

Second, you can make a back rotation, 1326 connects to deception can also be concatenated: take the crystal head network card one end: orange white, orange, green white, blue, blue white, green, brown white, brown, the other end does not pick up RJ-45 Directly connect 13, 26 directly, ie orange white, green, orange, other 6 wires don't want to pay --------------------- ---------------- 37, NT does not recognize the hard disk, only recognize the 8G hard disk solution with atapi.sys in SP6, atdisk.sys replaces the NT installation disk. Original file in the i386 directory, then run the installer or after the NT copying file to the hard disk, replace the files on the $ ** $ copy on the C drive with these two files, which can be on the large capacity hard drive Install NT4.0 without re-adjust partition ------------------------------------- -38, how do I use the netsh command to configure COPY to another IAS? Netsh aaaa show config> Path / file.txt netsh exec file.txt should be available in Netsh aaaa dump> file.txt in 2k3 ------------------- ---------------------- 39, can you install route and remote access components under Win2000 Pro to make it remote login server? 2K Pro is with RRAS The function, only the MMC Console, 2K Pro's RRAS, only one service, just start the management tool -> service Routing and Remote Access service (not started by default), then you can nearer Seeing "Incoming Connections", in the incoming connection, you can configure RAS customers, VPN customers, etc., 2K Pro only support three remote users, RAS, VPN, and direct connections to accept one customer, together client.

---------------------------------------- 40, Rras Console Displays An "Internal" routing InterfaceAfter you install Routing and Remote Access Services (RRAS), an interface named "Internal" appears under Routing Interfaces in the Routing and Remote Access Services management console.This interface is part of RRAS and represents all Remote Access Services (RAS) devices. All Ras Clients Are Part of this Interface. The "Internal" Interface Should Not Be deleted --------------------------------- -------- 41, how to increase the save time saved in the ARP buffer project? The time the default value saved in the ARP buffer project is two minutes, you can change it below: Open the registry editor; find hkey_local_machine / system / currentcontrolset / service / tcpip / parameters to find "Double-word value" under "New", After entering the name ArpCachelife, enter the number of seconds you wish to save; exit the editor, restart your computer. ------------------------------------- 42, modify the password of the domain controller? After using Erd Commander 2002, the ISO CD is started to change the DC local password. The DC is the hand to the hand. Use the local administrator password to log in to this machine (directory recovery mode), then configure the local security policy boot script, suppose you The domain administrator's account is admin, the content of the boot script should be the net user admin 12345678 saved as an admin.bat file, then restart. Then your DC's administrator password is changed to 12345678. Everything will be done. -------------------------------------- 43, to disable the Office XP text service? ? Removing CTFMon.exe from MSConfig does not disable text services. Because the text service will start automatically when you start a pro program. To prohibit text services, follow these steps: Step 1: Delete Chinese Optional User Enter Method For Windows 2000 systems, delete this feature: 1. Exit the Office program. 2. Click Start, click Setting, and then click Control Panel. 3. In the Control Panel, double-click Add / Remove Programs. 4. In the current installer list, select Microsoft Office XP products. Click to change. 5. In the Maintenance Mode dialog box, select Add or Remove Components, then click Next. The dialog box to select the installation option is displayed. 6. Click the plus number before the Office sharing function. 7. Click on the icon of the Chinese optional user input method to select not installed. 8. Click Update. Step 2: Remove the optional user input from the text service: 1. Click Start, click Setting, and then click Control Panel. 2. In the control panel, double-click the text service.

3. In the installed service list, except English (US), one by one deletes the remaining input method. Step 3: Run the Regsvr32 / U logout MSIMTF.DLL and MSCTF.DLL files. 1. Click Start -> Run. 2. In the Run dialog, type the following command: "Regsvr32.exe / u msimtf.dll" (excluding dual quotes) 3. Click OK. 4. Repeat 1 to 3 steps for MSCTF.dll files. After prohibiting the text service, you can add other input methods as before. -------------------------------------------------- -44, what is RPC? ? Inter-process communication (IPC) is a communication technology used by programs and processes run between multitasking operating systems or networking computers. There are two types of process communication (IPC). Local Process Call (LPC) LPC is used in multi-tasking operating systems such that tasks running simultaneously can be sessions. These tasks shared memory space allows tasks to synchronize and transmit information. Remote Process Call (RPC) RPC is similar to LPC, just working online. RPC begins to appear in a computer that runs a SUN microsystem and HP's running UNIX operating system. By IPC and RPC, the program can utilize other programs or computer processing processes. Client / Server Mode Calculation The remote process calls with other technologies such as messages, as a mechanism for inter-system communication. The client executes its own task, but provides backend file services by the server. RPC provides the client to provide a communication mechanism for the rear server application service, as shown in Figure R-4. If you want the client / server application to be a separate program, the server can run the data access section because it is recently separated by the data, and the client can run the data representation and the front end portion interact with the user. Thus, the remote process call can be seen as part of the part of the split program through the network. LPC is sometimes also known as coupling mechanism. Use this method to divide the program, do not need to copy the entire database or most of its programs to the user system when the user wants to access the data. In fact, the server only processes the request, even only performs some data calculations, and sends the resulting result to the user. Because when the data is placed in one place, the database synchronization is easy to implement, so many users can access the same data simultaneously. The distributed computing environment is a computer cluster connected by a communication system-network connection. It is easy to see this network as a computing platform. If it is equal, any computer can be a client or server. Some processing tasks can be divided into separate runners in parallel on different network computers, while independent programs are handed over to computer processing that best suits this task. This strategy can use computer idle resources to improve network efficiency. A typical enterprise network includes many heterogeneous computer systems that run different operating systems.

------------------------------------------------- 45 How to prohibit anonymous users from connecting your IPC $ sharing? We can implement key: hklm / system / currentcontrolset / control / lsaname: restrictanonymoustype: RONTRICTANONYMOUE: 1 | 2 Description: When you set this value to 1, anonymous users cannot enumerate the list of host users; set this value to 2 At 2, anonymous users cannot connect your IPS $ shares, not recommended 2, otherwise you can cause some of your services that cannot be started, such as SQL Server ...--------------- -------------------------------- 46, regedit.exe parameters: FileName Import .reg File Enter Registry / S Import .REG File Enter Registry (quiet mode) / e exported registry file example: regedit / e filename.reg hkey_local_machine / system / l: system specified system.dat / r: user specified User.dat / C compression [file name ] (Windows 98) ------------------------------------------------------------- --------- 47, ". REG" file Return Slightly Slightly Slightly, in order to make it easy to modify the registry, we will make some scripts, but no matter whether it is an INF file or a VBS script, I think it is still only .reg file is the most convenient. About the operation of the registry, common is to create, modify, delete. - Creating a creation is divided into two kinds, one is to create children (Subkey) Note: If you are not very clear about the naming of the registry, you can look at the Registry Naming Standard Manual (http://www.sometips.com/ Tips / RegistryHack / 204.htm) We create a file, the content is as follows: Windows Registry Editor Version 5.00 [HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM] Then execute the script, you have created a name "Test4adam" in HKEY_LOCAL_MACHINE / SOFTWARE / Children. The other is to create a project name that this file format is a typical file format, consistent with the file format you exported from the registry: Windows Registry Editor Version 5.00 [HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM] "Test1" = "Adam" "Test2" = HEX: 61 "Test3" = dword: 00000064Test1 type is "String Value" Test2 type is "binary value" TEST3 type is "dword value" Note: If you don't exist in your registry Test4adam This child, then the script will create the child for you. - Modifying modifications is relatively simple, just export you need to modify, then modify it with Notepad, then import it, here I will not repeat it.

- Delete let us first say to delete a project name, we create a file as follows: Windows Registry Editor Version 5.00 [HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM] "TEST1" = - Execute this script, HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM "test1 "It's deleted; let's take a look at the delete one child, we create a script with the following: Windows Registry Editor Version 5.00 [-HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM] Execute this script, HKEY_LOCAL_MACHINE / SOFTWARE / TEST4ADAM has been deleted. -------------------------------------------------- ------ 48, Power Users Group: Members of the Power Users group can modify the non-administrative user's account, start the default service without starting (note, a member of a Power User group cannot stop a system service) This group member cannot modify the administrator's account password, install the system service or modify the operating system file. -------------------------------------------------- ----- 49, Remote Backup and Restore: You can use the / saveboot: dsrepair command option in the Boot.ini file on the remote computer, boot the remote machine to enter the recovery mode, then use the planned task or The AT command implements backup and recovery. Implementing remote restarts You can use the corresponding management programs in 2000, XP, 2003.

-------------------------------------------------- ------ 50, RAS client to obtain DHCP options: the mobile users receive their IP configurations from the Remote Access Server, but they are not able to receive any DHCP options In order to enable this, a DHCP relay agent must be. configured on the Remote Access server. This will allow DHCPINFORM, which are used to obtain Windows Internet name Service (WINS) and Domain name System (DNS) addresses, domain name, Default Gateway or other DHCP options originating from the DHCP server, to reach The Mobile Clients. ---------------------------------------------------------------------------------------------- ---------- 51, use the netsh command to copy IAS settings: Remote Access Policies Are Not Stored In Active Directory; Thei store, THEIAS.MDB FILE. To Copy The IAS Configuration To Another Server Weme Must Type netsh aaaa show config /file.txt at the command prompt. This stores the configuration settings, including registry settings, in a text file. The path can be relative, absolute, or a UNC path. We must then copy the file we created to the destination computer, and at a command prompt on the destination computer, type netsh exec /file.txt. A message will appear indicating whether the update was successful or not .------ -------------------------------------------------- -52, DHCP User Class:

User classes allow DHCP clients to differentiate themselves by specifying a User Class option.When available for client use, this option includes a user-determined class ID that can help to group clients of similar configuration needs within a scope, such as providing a shorter lease time for portable computers that move frequently or use remote access often. Typically a DHCP server will be used to distribute different options that are specific to the needs of clients.To configure a User Class Lease Time we must open the DHCP console, select the DHCP Server, open Scope, Right click Scope options, select Configure Options, select the Advanced ban, Select appropriate Vendor Class and User Class (= Windows2000LapTopComputers in this example), Select 051 Lease, and Enter lease time: 14400 (4 hours = 14400 seconds ) ------------------------------------- -------- 53, Mobile DHCP Database: To Move The DHCP Database We Must First Stop The DHCP Service On The Old Server, Back Up The Registry Key HKLM / S . OFTWARE / Microsoft / DhcpServer / Configuration, and install DCHP on the new server.We must then stop the DHCP service on the new server and restore the Registry key from the old server onto the new server Then we must delete the contents of C: / WINNT / System32 / DHCP on the new server, copy the database file DHCP.MDB from the old server onto the new server but not the transaction logging (* .LOG) and checkpoint (* .CHK) files and start the DHCP Service on The New Server. Finally We Must Reconcile All Scopes on The New Server To Synchronize The Database with The Registry .---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ---------------------------- 54, DNS aging recording function:

Previous versions of Microsoft operating systems that do not support dynamic Domain Name System (DNS) require that a static DNS entry use a static IP address whenever possible. If we upgrade to Microsoft Windows 2000 and our present DNS server is Windows 2000, the IP address will remain the same, but the DNS "A" record remains static. However, the static PTR record is converted to a dynamic entry and is subject to the aging process. The Windows 2000 Dynamic Domain Name System (DDNS) client does not overwrite an Existing "a" Record if The IP Addresses Match. To Convert Static Entries To Dynamic Entries, We Must Use The / AgeallRecords Option In The DNSCMD.EXE COMMAND. ----------------- ---------------------------------------- 55, NBTSTAT A useful command: nbtstat CRR: Releases Names Registered with a Wins Server and The Renews Their Registrations. This Will Release Obsoleten Records and All Wins Clients Will Get Registered Properly Again. ------------------- -------------------------------------- 56, DNS f ast zone transfer format: By default, all Windows-based DNS servers use a fast zone transfer format, which uses compression and can include multiple records per TCP message during a connected transfer This format is also compatible with more recent Berkeley Internet Name Domain (. Bind -Based DNS Servers That Run Versions 4.9.4 and later. ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- --------------------- 57, 98/2000 / XP / 2003 Access XP problem Some people say that enable guest, some people even install NetBeui, etc. Since then, you will summarize the correct statement, pay attention to the 98/2000 / XP / 2003 access to XP / 2003 user authentication issues with normal network connection is normal and permission allowed. First of all, I can't access the guests why you cannot access. 1. By default, XP disables guest account 2. By default, XP local security policies are prohibited from accessing GUEST users from network.

By default, XP's local security policy - User Rights Assassin, "Empty Case Users can only log in to the console" are enabled, that is, any account of the empty password cannot be locally logged in, Guest The default empty password? Sometimes it will encounter another situation: When accessing XP, the username in the login dialog is gray, which is always a guest user, and cannot enter other user accounts. The reason is that this security policy is in the Strange (Management Tool - Local Security Policy - Security Options - "Network Access: Sharing and Security Mode for local accounts"). By default, XP's access method is the way "only guest", then you access it, of course, you cannot enter other user accounts for GUEST. So, accessing the easiest way to access the XP is: No need to enable guest, only modify the above security policy is "classic". Other systems access XP can enter account information. As for accessing 2003, by default 2003 disable guest, but there is no XP that hate the default contradictory guest mode sharing, so you can enter your username password access. -------------------------------------------------- ------ 58, IPX / SPX frame type:.. NetWare 3.11 uses the 802.3 frame type Netware 3.12 and above uses the 802.2 frame type This network has both NetWare 3.11 and NetWare 4.1 servers, so both Frame Types must be installed . Installation of multiple frame types on a Windows 2000 Professional requires editing of the Registry, specifically add both types to the multi-string value pktType in HKLM / SYSTEM / CurrentControlSet / Services / NwlnkIPX / Parameters / Adapters / , where Is The network adapter Identifier. --------------------------------------------------------------------------------------------------------------------- ------------- 59, four coding methods: ANSI: System preset standard text storage format. ANSI is an abbreviation for American National Standards Institute. It was established in 1918 and a voluntary organization with more than 1300 members, including all large computer companies. ANSI is specifically established for computer industry, which is a considerable standard of standards in the world. Unicode: The integration of all major directive files in the world, including the public segments used in business and personal computers. When you use the Unicode format to store files, you can use the Unicode Control Character Assistment Description Language Coverage, such as Arabic, Hebrew. User Enter the text containing Unicode characters in "Notepad" and stores files, the system will prompt you to select Unicode encoding in Save As, these characters will not be lost. You need to remind everyone that some Windows 2000 fonts cannot display all Unicode characters.

If some characters are missing in the discovery file, simply change it to other fonts. Unicode Big Endian: The text bitmap (storage unit) in the Unicode file established on the BIG-Endian processor (such as Apple Macintosh) is arranged in order of the text bit tuplement of files established on the Intel processor in contrast. The most important bit group has the lowest address and will first store a larger end of the text. In order to allow users of such computers to access your files, select the Unicode Big-Endian format. UTF-8: UTF means universal character set transformation format, UTF-8 is the 8-bit form format of Unicode. If you use only important information of 8 bitmines within the same location in group, you can choose the UTF-8 format. -------------------------------------------------- ----- 60 CA certificate revocation:. revoking a certificate is a two step process first we must revoke the certificate and then create (this is done automatically) and publish the certificate revocation List (CRL) Finally, the certificate revocation Lists (CRLS) Must Be Accessible To All Users. It Should Be Put in a Network Share and the Users Should Have Appropriate (Read) Permission to the Share. ----------------- -------------------------------------- 61, hide your computer in the network neighbors, 2000 There are two ways to use: a. Editing Registry: Setting HKEY_LOCAL_MACHINE / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / CURRENTCONTROLSET / SYSTEM / LANMANSERVER / Parameters Set to 1 from 0 to 1, this value is DWORD type .b. Run as follows: Net Config Server / Hidden: YES This user may connect to your computer but cannot see you in the network neighbor. Microsoft Cluster Solution

Excerpted from the article on the WINDOWS 2000 magazine

For years, Microsoft has always been dedicated to expanding the telescopic capacity, usability and reliability of your own server solutions. Clustering technology has proven to be an effective way to achieve this goal, and Microsoft fully supports cluster technical philosophy and is committed to making it part of the Microsoft operating system and related products. With the formal release of Windows 2000, Microsoft Cluster Solutions have gradually entered the maturity phase. Scalable ability, availability and reliability clustering is a set of independent computers that run the same application through synergies and provide a single system for clients and applications. The goal of clustering technology is to further improve telescopic ability, availability and reliability through multi-layer network structure. Scalability refers to the ability to process continuously improved workloads under the premise of maintaining acceptable performance. Hardware devices scalability (in-contained expansion, statement, depending on Microsoft), depends on various operations in large computers with extension capabilities. Software product scalability (epitaxial expansion) depends on the cluster formed by organizing multiple computers together by collaborative working mode, which is two different concepts with RAID drive arrays. In fact, Microsoft uses a non-regular term computer redundant array (RAC) to refer to its own extension expansion cluster. As you can improve performance by adding a disk to the RAID array, you can also improve performance by adding nodes to an extensive expansion cluster. Availability and reliability are two closely related but slightly different concepts. Availability refers to the quality, alternate, and easy accessibility, and accessibility. Reliability refers to the degree of firmness of the system. Even if the most reliable system will always have problems. Hardware Equipment Manufacturers offer preparation measures in advance in advance in a presence of possible failures in advance in key technologies such as disk drives, power supply, network controllers, and cooling fans. However, the redundancy provided on a computer cannot avoid the application failure for the user. If the database software on a server fails, even though this server may be very reliable, the functionality provided by software and servers will not be available. It can be seen that a single computer is not unable to compete for all necessary scalability, availability and reliability challenges that are borne by the cluster. The cluster can simulate the RAID array to provide usability and reliability. In a fault-tolerant disk configuration such as RAID 1 or RAID 5, all disks work together in redundant array. If a piece of disk is faulty, you can unplug it and insert a new disk, and the rest of the array will continue to run - no configuration, no need to install, the most important thing is, no downtime. The RAID system automatically reconstructs the new drive to make it work together with other drives. Similar to this, when a computer in the cluster fails, you only need to replace it with a new system, and the entire cluster will continue to operate. Some cluster software can automatically configure the server and integrated into the cluster - all related operations are done in the cluster in available. Four cluster solutions Microsoft provides four basic clusters: Microsoft Cluster Services (MSCs), Network Load Balancing (NLB), Component Load Balancing (CLB), and Application Center 2000. These services are available through three solutions: MSCS, NLB and Application Center. The CLB is one of the components of the Application Center and can only be applied by Application Center. NLB can be used both by Application Center or as an independent solution. Windows 2000 Advanced Server and Windows 2000 DataCenter Server itself contains MSCs and NLB, but you must purchase Application Center separately.

Table 1 summarizes the availability of these four cluster technologies in different Windows 2000 Server and Windows NT Server 4.0 products. As you think, there is no such technology for Windows 2000 Professional or Windows NT Workstation 4.0. Table 2 lists some cluster technical features. You can refer to this form when conducting technical comparisons in the rear of this article. MSCSNLBCLBAPPLICATION CENTERWIN2K Server Unavailable support (need to use AppCenter support (need to use third-party IP load balancing tools) Win2k AS contains (up to 2 nodes) to include support (need to use AppCenter) support (using NLB) Datacenter (up to 4 nodes) Contains support (need to use AppCenter) Support (using NLB) NT Server 4.0 Unavailable Support (called WLBS) Unavailable NT Server 4.0, Enterprise Edition contains (up to 2 nodes) Support (called WLBS) Unavailable table 2 Cluster technical characteristics

MSCSNLBCLBAPPLICATION CENTER Purpose Application Failure Recovery and Fault Return IP Communication Load Balancing COM Object Load Balancing Create and Manage Web Areas Advanceability and Manageable Availability and Scalability Availability and Scalability Availability, Scalability and Manageable Benefits Each The maximum number of nodes in the cluster (for Win2K AS) or 4 (for Datacenter) 32 16 16 Cluster type shared storage mechanism No shared resource No resource No resource status information is stateless (if needed, you can support There is a stateless connection) No status No status requires modification of the server application requires no need to do not need to use a dedicated hardware device requires no need to independence, whether it is independent, whether it is not required (need using appCenter) is the first microsoft cluster service Codentname WolfPack and the MSCs called Microsoft Cluster Server and Microsoft Cluster Servers is the first heavy boxing in Microsoft in the NT cluster technology. It is a recognized best Microsoft cluster solution. In the MSCS cluster, MSCS software can establish a connection to a physical computer running on a high-speed network. Typically, the computer in the cluster can share the same storage subsystem and function in the "activity-activity" manner, which means that all cluster computers (nodes) can be actively collaborative to work together, and in a certain The node fails to share its work. Figure 1 shows a 4-node MSCS cluster. Figure 1 The main purpose of the 4-node cluster MSCS implemented by Windows 2000 MSCs is to improve application availability through its fault tolerance. Fault tolerance refers to the clustering function that transplays the relevant processing procedure from a fault application on a node (due to hardware device failure or software errors). When the fault application is restored, the cluster should be able to implement "fault return" for the original cluster node. MSCs can fail to recover the applications running on the cluster without losing any fault application related data, and can maintain user and application status during the fault recovery process. This type of cluster function is referred to as a status cluster function. In contrast, NLB, CLB, and Application Center provide stateless cluster functionality and dynamic load balancing capabilities while enhancing availability (I will discuss these content in a later portion). For applications such as email servers, database applications, MSCs is a good way of operation. Suppose you decide on a 4-node MSCS cluster running Microsoft Exchange 2000 Server. When you install MSCS software and apply the Exchange 2000 version for cluster, you can configure the cluster to make Exchange 2000 failure on the backup node when the primary node fails. When the fault occurs, there is certainly a user session in an open state on the primary server, however, MSCs can quickly complete fault recovery without losing any data. Backup nodes will replace workloads and related data from the fault node, and continue to serve users. MSCS also allows users to continue working during the application upgrade process. You can take scrolling upgrade (for example, each time a cluster node is up to the application and make sure the application on other nodes continues to be available) without having to stop using the application during the upgrade process. For example, suppose you have a two-node cluster. Where node 1 runs Exchange 2000, Node 2 Run Microsoft SQL Server, you want to configure this cluster so that Exchange 2000 and SQL Server can recover each other in the necessary cases.

When you need to upgrade SQL Server, you can start SQL Server failure recovery capabilities on Node 2 via MSCS Cluster Manager. When the node 1 is running tasks in the SQL Server (while continuing to run Exchange 2000), you can upgrade SQL Server on Node 2. When the upgrade work is completed, you can run the SQL Server from the node 1 to the node 2 by fault return, and repeat the same software upgrade operation on the SQL Server on Node 1. When the node 1 is also upgraded, the entire SQL Server software completes the upgrade task without affecting the user. Unlike other three Microsoft cluster solutions, you usually not use MSCS to expand your application for more users. The MSCS cluster cannot provide dynamic load balancing capabilities between nodes in nodes, or implements application distribution operations as in stateless non-shared modes. In fact, the only feasible way to implement application scaling capacity using MSCs is to manually assign applications to different cluster resources during the installation process. For example, if you need to serve 5000 users on the Exchange 2000 platform, you can apply 2-node activity-active clusters and serve 2,500 users at each node. In this way, you can not only get the performance advantages achieved by the two servers to provide services to users, but also implement the necessary availability in the absence of failure. However, when the failure recovery occurs, the remaining nodes must be able to provide services to all 5,000 users before the fault node is recovered. Network Load Balancing NLB referred to as a Windows NT Load Balancing Service (WLBS) can be reasonably allocated on the node of multiple running NLB software. NLB provides scalability and availability for an IP protocol, such as a web server. As users have increasing the demand for server resources, NLB allows you to add servers to handle workloads. For example, the load pressure of the Exchange 2000 server is significantly relieved by using NLB implementing Microsoft IIS communication front ends facing Outlook Web Access (OWA). The NLB cluster will request the client to route to one or more backend servers. If an NLB node has a shutdown failure, other nodes will automatically assume additional loads left by them, and the user will not feel any service interrupt. NLB underlying software is a network device interface specification (NDIS) driver between NIC and TCP / IP. You should install this driver on each server in the NLB cluster. All NLB nodes share the virtual IP address of the same representative of the network resources (such as a web server). All NLB servers monitors user requests, but only one server responds to these user requests. Load Balancing Architecture Based on Fast Hash Algorithm is responsible for merging client IP addresses and port numbers and determines which server responds. You can specify a similarity rule to be able to assign different loads on different servers (for example, you can specify some servers to get more than other server workloads). Through a heartbeat feature, all NLB nodes can change in time (such as new nodes or node failures). When the cluster changes, NLB will initiate a gathering process to automatically coordinate changes in the cluster and re-allocate the workload of the entering system in a transparent manner. The original origin of NLB dates back to the 1998 Microsoft Acquisition Headquarters at the Valence Research in Oregon. The CONVOY cluster software developed by Valence is later WLBS (NT Server 4.0 and a plug-in product in NT Server 4.0 Enterprise).

In Windows 2000, although Microsoft further enhances the functionality of WLBS and renaming it, its core technology has not changed. NLB is an integrated component in Windows 2000 Advanced Server and DataCenter network services. As long as you run on a separate computer, the MSCs and NLBs in Windows 2000 can be well implemented well - the configuration scheme shown in Figure 2, please refer to Figure 2. Since there is a potential hardware device sharing conflict between MSCS and NLB, Microsoft does not recommend running MSCS and NLB on the same computer. For more information on MSCS and NLB installation information, check the Microsoft Article of the topic of Windows 2000 interaction between MSCS and NLB. Figure 2 The N-Node NLB Cluster Component Load Balancing in a single virtual IP server mode is a new thing for Windows 2000, CLB is a new thing. Similarly, for Windows 2000, COM as next-generation COM technology is also a new thing. Com combined with COM, Microsoft Server (MTS), and related system services, making Windows 2000 into a design, development, deployment and maintenance of a component-based excellent platform. Briefly, COM is a COM technology that includes a series of related system services (including services that allow you to distribute components between multiple systems). One of the COM services enables load balancing access to COM objects. The CLB is only a load balance cluster function - that is, the workload generated by multiple servers sharing activation and execution of COM objects. Similar to NLB, the need for CLB is also derived from availability and scalability. When you run critical applications composed of a COM object, the application or server failure will result in serious consequences. The CLB ensures that the application continues to run in the event of a failure, and avoiding the user feels that the service performance has declined. In addition, certain COM objects are quite large and fairly complicated, running these COM objects on a single server and other key applications such as IIS will seriously affect system performance. To provide scalable ability in this case, you can remove COM objects from the IIS server and reasonably distribute them on multiple servers in your own CLB cluster. Suppose you are a computer manufacturer with your business web site, your business site is responsible for providing users with product information, product technical support, and online ordering services. Close to customers around the world use your product 24 hours a day, so your Web site must be available at any time and maintain a good running state. To do this, you can run NLB on the way shown in Figure 2 in the WEB server with the backend MSCS database cluster access capability web server. However, if the service you provide uses a large amount of implementation logic based on COM object coding, although these objects can be run on the web server, because the computer running the web server must also process COM objects, therefore, Web The response time of the server will be very slow. At this point, you may need to use the CLB. Figure 3 depicts how to deploy CLB clusters on a Web site with high availability and scaling capacity. The CLB can implement load balancing for business logic for providing COM objects in the application intermediate layer. (CLB cluster requires Application Center to provide implied support, we will explain the relevant information in detail in the "Application Center" section, now you only need to know why you need to use the CLB.

Figure 3 Implementing the COM Object Load Balancing Access Method CLB to determine the next request by using the server response time with the cyclic algorithm. The CLB performs votes in the cluster in a pre-set time interval to determine the server response speed (the server response time directly reflects their busyness). The CLB is sorted by the response time, and the server with the shortest response time is in the header and is responsible for processing the next COM activation request. Thereafter, the CLB will work in the list in the list before the next voter is separated. Since all of the above processing work is completed in real time on the network, you can see that if you add CLBs on a slower or more crowded network, it is possible to cause network competition issues. Therefore, you should deploy a CLB cluster on a high-speed trunk network with at least 100 Mbps transfer rates. Typically, you should not deploy CLB clusters on a regular enterprise network that carries other network communication. The way to distribute COM objects in the CLB cluster is not applicable to all situations, and you must make decisions using CLB on the basis of application requirements. The cluster function adds loads related to client requests throughout the network and to select a server to satisfy the client request and activate the COM object. In some cases, since the application only uses a small amount of lightweight COM objects, a simple way to instantiate objects locally in the web server may provide higher performance. It is necessary to consider the use of CLB applications including:

Compared to comparative "complicated" COM objects that make up business logic and must be run on the high speed server. Safety is one of the main concerns, and you want to independent COM object by placing it in the back of the additional firewall. For development or design reasons, your COM application is divided into multiple hierarchies, and you need to use CLB at separate hierarchies. CLBs are not available in all Windows 2000 Server product family members, and you cannot purchase CLBs through separate products. Initially, Microsoft has intentionally contains CLBs in the Windows 2000 Server product family, but in September 1999, Microsoft separated it from Windows 2000 Release Candidate 2 (RC2) and put it into a kind of Application Server. In the new product, the only way to get the CLB is to use Application Server. Application Center Application Center is one of the components of the Microsoft .Net Enterprise Server product family, its predecessor is the Windows Distributed Internet Applications (Windows DNA) server. The main purpose of Application Center is to implement a single management point for the Web area (such as a multi-physical web server that provides the same web content and collaboration with each other), providing a unified user interface (UI) and utilizing NLB and CLB to implement load balancing. By using Application Center, you can create a new cluster, join an existing cluster, add or remove cluster members, deploy new content, configure load balancing, and monitor cluster performance. The final result is to provide telescopic capacity, ease of use, and high reliability for the web area, and make it a single web server in the eyes of external users. As more and more key applications begin to close to the web, the importance of these functional features is also growing. For complete Application Center topologies, all Microsoft cluster technology works with them, please review Figure 3 again. The NLB cluster can be a cluster composed of IIS servers, and the CLB cluster can be used to provide business logic. Both of the two constitute an Application Center web cluster, while the database cluster can be implemented by MSCS. Suppose you have an e-commerce site and is planning for large-scale products scrolling upgrades. During this period, many customers want to purchase your products through this site. This situation will significantly increase the Web site load, but you cannot determine how big this increase is. You can add new servers as needed, but the settings for these servers are very complicated. For product scrolling upgrades, you can easily extend Web site performance like adding a server in the RAID array to add a server to the Web area. This application situation is the case where the RAC concept is applied. Application Center provides a wizard program to create a cluster and add a new server to the cluster to deploy new content and configure the cluster member. When you create a new cluster, you need to define a cluster controller, which is not only a cluster member, but also maintains all configuration information. After that, you can specify other members in the cluster. When you complete the above work, Application Center deploys COM settings, CryptoAPI settings, registration table, and Windows Management facilities (WMI) settings, file system information, II sub-database settings, and web server contents for each new cluster member.

In the end, you will get a cloning cluster, and you can easily add or delete cluster members using the Application Center Manager. In addition, Application Center will also complete monotonous NLB configuration and deployment through a transparent manner. Application Center will also support other third-party IP load balancing tools other than NLB. When writing this article, Microsoft is providing support for Cisco System's LocalDirector, F5 Network's BIG-IP, and ALTEON Websystem's AceDirector. However, Application Center does not integrate management operations for these load balancing tools as integrated NLB management, so you need to maintain these third-party load balancing tools by performing additional work. You can install the full version of the necessary components that you need to create an Application Center cluster on the following version of the Windows 2000 operating system platform: Win2k Server Service Pack 1 (SP1) Win2k AS SP1 DATACENTER SP1 is to support Application Center, you must have Win2k sp1. Otherwise, you will only install Application Center Manager for remote management of Application Center and IIS. The following version of the Windows 2000 and Windows NT operating system supports Application Center Manager:

Win2k Professional Win2k Server Win2k AS Datacenter NT Workstation 4.0 SP6 or later (using X86 processor) NT Server 4.0 SP6 or higher (using X86 processor) When you read this article, you can pass Microsoft's new type. The way the .NET Enterprise Server product is independently purchased Application Center. Cluster technology and Microsoft .NET You may want to know how Microsoft cluster solutions adapt to the company's future business development needs. The Windows service has been awarded the new name of Microsoft .NET. Figure 4 compares Windows 2000 and Windows DNA to Windows .NET and .NET platforms. Microsoft Cluster Solutions are available in two Microsoft .NET areas: Windows .NET (MSCs), and .NET Enterprise Server (NLB, CLB, Application Center). Windows .nt represents the evolution process of the Windows 2000 operating system, the next two version of the code is Whistler and Blackcomb, respectively. .NET Enterprise Servers contains the following products (most of them will be included in BackOffice and have integrated XML support capabilities):

Exchange 2000 SQL Server 2000 BizTalk Server 2000 Application Center 2000 Host Integration Server 2000 Commerce Server 2000 Internet Security and Acceleration (ISA) Server 2000 Windows DNA from 4 to Microsoft .NET platform evolution Windows.NET and .NET Enterprise Servers are Microsoft Perfect and enhance both basic components during the forward process of the .NET platform. Microsoft attaches great importance to WINDOWS scalability, availability and reliability. These features are also more important in Microsoft as a business platform to enter the Internet. Obviously, the emergence of powerful cluster technology such as MSCS, NLB, CLB, and Application Center has a crucial role in Microsoft's key business application support. I want to see these cluster technology - especially the application center, and Microsoft's new .NET technology, such as active server page (ASP ) loaded in early Professional Developer Conference, is further enhanced and expanded. Immediately take up some time to understand the working principle of these powerful cluster solutions, and they may help you solve. As the procedural executive of the BMC Software's Patrol Business Unit, Greg Todd is responsible for developing the development direction of Windows 2000 solution technology for the company. He began working with Windows NT and related technologies since 1993. If you want to get in touch with him, please call: gregt@bmc.com.

Related Microsoft Web Site "Exploring Windows Cluster Technology" http://www.microsoft.com/windows2000/guide/server/features/clustering.asp "Network Load Balancing Technology Overview" http://www.microsoft.com/technet/ Win2000 / nlbovw.asp http://www.microsoft.com/windows2000/library/howitworks/cluster/nlb.asp "Windows 2000 and Component Load Balancing" http://www.microsoft.com/windows2000/news/fromms/ CLB.ASP "Microsoft Application Center 2000" http://go.microsoft.com/fwlink/?Linkid=591 "Windows 2000 Advanced Server" http://www.microsoft.com/windows2000/guide/server/solutions/ overview / advanced.asp "Microsoft Windows 2000 Datacenter Server" http://www.microsoft.com/windows2000/guide/datacenter/overview/default.asp "Windows DNA Application Services" http://www.microsoft.com/net / "Microsoft Windows DNA" http://www.microsoft.com/dna/default.asp ".Net platform and Windows DNA evolution process" http://www.microsoft.com/business/products/webplatform/evolution.asp "Microsoft .NET" http://www.microsoft.com/net/ "Microsoft PDC 2000" http://msdn.microsoft.com/events/pdc/ MICORSOFT hopes that the information provided herein can Enough to help you. In the process of using this article, you will assume all your own behavior. All the information in this article is provided in the form of "not guaranteed". Microsoft does not make any clear or implicit guarantees for specific purposes, title or non-infringement. All mentioned in this article Third-party products or information has not been licensed, recommended, supported or guaranteed by Microsoft. Microsoft does not assume any direct, indirect, special, attached or corresponding legal liability for the damage you caused by using this information, even considering the possibility of this damage. When there is a change in the prices mentioned herein, you will not be notified. LINUX configuration and optimization experience

As a Linux enthusiast, the author has installed various Linux issuings. During each installation process, a large number of configuration optimization works a lot of time and is easy to cause some detail missing. This article is in the case of installing and using RedHat 6.1, talking about some of your own experience. About the hard disk partition author is to establish a separate partition when installing Linux, and is used to store user private data, and can be used for the configuration file for the backup system. & nbs p; Editing / etc / inittab file Most of the Linux issued suite created six virtual console, actually three already enough, and saving valuable memory space. Edit the / etc / inittab file, plus a # in front of the following three lines. 4: 2345: Respawn: / sbin / mingetty TTY4 5: 2345: Respawn: / sbin / mingetty Tty5 6: 2345: Respawn: / sbin / mingetty TTY6 then execute the init Q, reread the system re-read / etc / inittab file, close TTY4, TTY5, TTY6 Virtual Console. If you don't consider the security of the system, you can use the fourth virtual console to view a variety of system information, edit the /etc/syslog.conf file, join:. ?? / dev / tty4 Finally, execute killall -hup Syslog. Editing the / etc / inputrc file As long as you edit the / etc / inputrc file, set the INPUTRC's environment variable, use the command completion of the Tab key to make the Doskey functionality similar to MSDOS. This is the same for various operations of the document and the directory. Edit /etc/rc.local file If you need to customize your login information, pay attention to modify the /etc/rc.d/rc.local file, because each execution /etc/rc.d/rc.local file, you need to modify / etc / issue.net file, we can edit your own / etc / Issue file, speed up the Linux boot process, and add the following: echo "base = 0xD8000000 size = 0x800000 type = write-combining"> > / proc / mtrr # enable numlock for tty in / dev / tty [1-3]; do setleds -d Num <$ TTY DONE Custom VIM Environment In Linux Environment, there are many editors, mainly Emacs, Vim , Joe, etc., for those who have never used UNIX or Linux, the author's suggestion is best to find a few related reference books, first understand these editors, so after installing Linux, editing and viewing some configuration When the document is, it will not know what it is. In general, you can install both Windows 98 and Linux in your own machine. By installing the MTOOLS package, you can use the corresponding file system, allowing in MSDOS, Windows 98 with Linux. Between file systems, implement file read, write, move, and display. Suppose Windows 98 is installed under / dev / hda1, / dev / hda5 partition, modify the /etc/mstools.conf file, add the following: Drive C: file = "/ dev / hda1" drive d: file = "/ dev / HDA5 "This can be executed in MDIR C:.

Copy various configuration files according to the purpose of the machine, edit the copied corresponding profile, such as Samba profile /etc/smb.conf, domain name server configuration file, DHCP profile /etc/dhcpd.conf, etc. Under normal circumstances, once these files are configured, they rarely change, which can save a lot of time. Note that the original configuration file must be backed up, if the configuration file is changed, you have to make a backup! Compiling Linux kernels Generally, the kernels installed by various Linux system suite are suitable for most machines, but the disadvantage is that there are many modules that do not need to be at all. Only the kernel is configured according to your own needs and the machine hardware conditions can be achieved. Note Before compiling the kernel, you must have a comprehensive understanding of your machine, so you can do it in your heart. The following is a brief step of compiling the kernel: (Take RedHat6.1 as an example) 1. Go to / usr / src / linux, edit the makefile file, modify extraversion = -xx, where XX means compiling version number, it is best to use numbers, And greater than the original value. 2. Select the corresponding option to configure the kernel according to the hardware configuration of your machine. #Make menuconfig 3. Relock the dependency of each file and clean up the previously generated target files and other files. #Make clean; make deposit 4. Compile the kernel and installation and install the module. #Make bzimage #make modules; Make Modules_Install 5. Install the kernel. # Cp arch / i386 / boot / bzimage /boot/vmlinuz-y.y-xx copies the kernel to the / boot directory, where y.yy represents the Linux version number, XX represents the version number compiled. # CP system.map /boot/system.map-yyyy-xx # cd / boot # ln-s system.map-yyyy-xx system.map Finally, edit LILO's configuration file /etc/lilo.conf . 6. Test the new kernel. After the boot is completed, pay attention to check the /lib/modules/y.y.yy-xx directory, create a Modules.Dep file. If the file exists, you can modify /etc/rc.d/rc.sinit, and the contents of the command to execute the module dependencies are removed, so that the LINUX startup process can be accelerated. Note that before the new kernel is tested, the user is best to keep the original kernel so if the new kernel boots Linux is unsuccessful, the old kernel can be used to boot the Linux operating system. User management of Win2000 Advanced Server

In the LAN in Windows 2000 Advanced Server, the objects in the domain, especially for the user, especially for the user. This article is to introduce the user-managed "three-axis ax" of Windows 2000 Advanced Server, namely: profile, folder, and management template.

Profile

The settings of the user's desktop, displays, networks, printers, and mouse are managed by user profile. The configuration file is divided into the following three categories:

1. Local user profiles, usually in the "C: \Documents and settings \% username%" path, users can assign, not in control.

2. Roaming the user profile, which is set to the network domain login user, which is generally located in a shared folder of the server, ensuring that users can get their favorite user environments regardless of which PCs login from the domain.

figure 1

3. Force user profiles, this is a strengthening of roaming user profiles. The setting method is: Enter the Administrative Tools \ Active Directory User and Computer, enter the UNC path name such as "\\ Server \Share \% Username" in a "Privileger" of a user account (Figure 1). Of course, the shared folder should be built in advance. When you log in with this account, you will see the file named "NTUSER.DAT" in that path, which is a bit shy (Hidden property). They are very different, and they can be changed when roaming. It can be saved; and forced must change the above files to "NTUSER.MAN", and although the login environment can be modified, only for this use, can not save, The next login is still "outside the lantern - still". Folder folder

Folder is rendered in two steps.

1, GPO settings

Windows 2000 Advanced Server provides four items for redirect control, namely: Application Data, My Documents, Desktop, and Start Menu four folders, you can set them separately. Taking the Redirect "Start Menu" as an example, the operation step is: Enter the Administrative Tools \ Active Directory User and Computer ", select" OU \ Group Policy \ Edit \ User Configuration \Windows Configuration \ File Heavy Direction ", figure 2). Right click on "Properties", see "Target" and "Settings" two tab items. Set the "Target" item first, in the Set box, generally select "Basic", and "Target Folder Position" is set to "\\ Server \Share". "Setting the" Tab Item best "Delete the policy back to the local user profile location".

figure 2

2, shared folder security settings

Previous step only indicates the position of the folder, but does not perform restrictive settings, to do this, you must start from NTFS security options. Find the corresponding redirect shared folder, enter "attribute \ security", only give "Everyone" with "read only" power, and give the network management with the "full control" power, this, each domain user's "start menu "Setting the right to escape your" like Buddha's palm "(as shown in Figure 3).

image 3

Management template

With the management template (Figure 4), "Documents", "Shutdown", "Run", "Search", etc. on you, using the management template (Figure 4), "Start" menu. "Desktop" items can also hide, banned, all icons on the "destroy" desktop. Some readers will say that there is a registry from Windows 95, which can do the above modification settings. But the intuitiveness of the management template is non-registration performance. Moreover, the registry is only for a user, and the management template is OU unit (an OU can put N users).

Figure 4

This "three ax" can be described as an ax ax. However, Tang Yan's tightening is not old. The settings of "My Documents" in the folder are constant, which is intended to make it easy for users to master their own documents anytime, anywhere. If you are hard to give it a "only reading", isn't it to carry it with the origin? In addition, if some people do have appropriate autonomy because of the work needs, you should also "open".

Win2000 Command Complete Works

accwiz.exe> ​​Accessibility Wizard for walking you through setting up your machine for your mobility needs. Accessibility Wizard acsetups.exe> ​​ACS setup DCOM server executable actmovie.exe> ​​Direct Show setup tool directly shows the installation tool append.exe> ​​Allows programs to Open Data In Specified Directories As IF The CURRENT DIRECTORY. Allows the program to open the data amiable in the set of data ARP.EXE> Network Display and Modify IP - Hardware Addresses displays and change the computer's IP and hardware physical address of the corresponding list at.exe > AT is a scheduling utility also included with UNIX run a task atmadm.exe> ​​displays statistics for ATM call manager. ATM call Manager statistics attrib.exe> ​​display and modify attributes for files and folders display and change file and folder attributes autochk .exe> ​​Used to check and repair Windows file systems to check and repair the file system autoconv.exe> ​​Automates the file system conversion during reboots automatically during the boot process transformation system autofmt.exe> ​​Automates the file format process during reboots format during startup Process Autolfn.exe> ​​Used for Formatting Long File Names Using long file name Format Bootok.exe> ​​Boot Acceptance Application For Registry Bootvrfy.exe> ​​Bootvrfy.exe, A Pr ogram included in Windows 2000 that notifies the system that startup was successful. Bootvrfy.exe can be run on a local or remote computer. Bulletin started successfully cacls.exe> ​​Displays or modifies access control lists (ACLs) of files. display and edit ACL Calc.exe> ​​Windows Calculators Calculator CDPlayer.exe> ​​Windows CD Player CD Player Change.exe> ​​Change {User | Port | Logon} CHARMAP.EXE> Character Map Character Map CHGLOGON.EXE> SAME AS Using "Change Logon" Start or deactivate session logging chgport.exe> ​​Same as sale "change port" change port (Terminal Services) Chgusr.exe> ​​Same As Using "Change User" Change User (Terminal Services) Chkdsk.exe>

Check the hard disk for errors similar to Scandisk 3 Stages must specify a Drive Letter disk testing procedures chkntfs.exe> ​​Same as using chkdsk but for NTFS NTFS disk testing procedures cidaemon.exe> ​​Component of Ci Filer Service Component of Ci Document Services cipher.exe > Displays or Alters The Encryption of Directories [Files] ON NTFS Partitions. Displayed or changing encrypted files or directory CISVC.EXE> Content Index - It's The Content Indexing Service for i Index CKCNV.EXE> Cookie Convertor Transformation Cookie cleanmgr.exe> ​​disk cleanup, popular with Windows 98 disk cleanup cliconfg.exe> ​​SQL Server client network Utility SQL client network tools clipbrd.exe> ​​Clipboard viewer for Local will allow you to connect to other clipboards Clipboard Viewer clipsrv.exe > Start the clipboard Server running Clipboard services clspack.exe> ​​CLSPACK used to create a file listing to establish a list of the system files of system packages list cluster.exe> ​​display a cluster in a domain displays the cluster _cmd_.exe domain> Famous command prompt nothing Something! CMDL32.EXE> Connection Manager Auto-Download Auto Download Connection Management CMMGR32.EXE> Connection Manager Connection Manager CMMON32.EXE> Connection Manager Monitor Connection Manager Monitor CMSTP.EXE> Connection Manager Profile Manager Connection Treator Configuration File Setup Comclust. EXE> About Cluster Server Cluster Comp.exe> ​​Comcluit Add, Remove, or Join A Cluster. Compare two files and file sets * Compact.exe> ​​Displays or alters the compression of files on ntfs partitions. Display or change NTFS partition Compressed status of files Conime.exe> ​​Console IME IME Control Control.exe> ​​Starts The Control Panel Control Panel Convert.exe> ​​Convert File System To NTFS Conversion File System to NTFS Convlog.exe> ​​Converts MS IIS Log Files Convert IIS Log File format to NCSA format CPROFILE.EXE> Copy Profiles Conversion Mode Cscript.exe> ​​MS Windows Scripts Host Version 5.1 Compared with this host version CSRSS.EXE> Client Server runtime process client server runtime process csvde.exe>

Comma Separated Variable Import / Export Utility to format conversion program dbgtrace.exe> ​​Terminal Server and related dcomcnfg.exe> ​​Display the current DCOM configuration. DCOM configuration properties dcphelp.exe>? Dcpromo.exe> ​​Promote a domain controller to ADSI AD installation Wizard Ddeshare.exe> ​​Display Dde Shares On Local or Remote Computer DDE Share DDMPRXY.EXE> Debug.exe> ​​Runs Debug, a Program Testing and Editing Tool. Is Debug! dfrgfat.exe> ​​Defrag FAT file system FAT partition Disk Defragmenter dfrgntfs.exe> ​​Defrag NTFS file system NTFS partition Disk Defragmenter dfs_cmd_.exe> ​​configures a Dfs tree configuration of a DFS tree dfsinit.exe> ​​Distributed File System Initialization distributed file system initialization dfssvc.exe> ​​distributed file system server distributed file system server diantz.exe> ​​MS Cabinet Maker make CAB file diskperf.exe> ​​Starts physical disk performance counters disk performance counters dllhost.exe> ​​dllhost is used on all versions of Windows 2000. DLLHOST IS The Hedost Process for All Com Applications. Main Process of All COM Applications DLLHST3G.EXE> DMAdmin.exe> ​​Disk Manager Service Disk Management Services DMREMOTE.EXE> Part of Disk Management Disk Management Services Part of DNS.exe> DNS Applications DNS doskey.exe> ​​recalls Windows command lines and creates macros command line to create a macro dosx.exe> ​​DOS Extender DOS extended dplaysvr.exe> ​​direct Play Helper run directly help drwatson.exe> ​​Dr Watson for 2000 Fault Detector Dr. Watson error Detection DRWTSN32.EXE> Dr Watson for 2000 Viewer and Configuration Manager Washing Doctors Show and Configuration Management DTCSetup.exe> ​​Installs MDTC DVDPLAY.EXE> Windows 2000 DVD Player DVD Play DXDI Ag.exe> ​​Direct-x Diagnostics Direct-x Diagnostic Tool EDLIN.EXE> Line-Oriented Text Editor. The text editor of the command line! Edlin.exe> ​​Line-Oriented Text Editor. The text editor of the command line (history is long!) Esentutl.exe> ​​MS Database Utility MS Database Tool Eudufactdit.exe>

Private character editor Ture Type Private Character Editor eventvwr.exe> ​​Windows 2000 Event Viewer Event Viewer evnt_cmd_.exe> ​​Event to trap translator; Configuration tool evntwin.exe> ​​Event to trap translator setup exe2bin.exe> ​​Converts EXE to binary format converter EXE file to binary expand.exe> ​​Expand files been compressed solution have that compression extrac32.exe> ​​CAB file extraction utility solution CAB tool fastopen.exe> ​​Fastopen tracks the location of files on a hard disk and stores the information in memory for fast access. Quickly access the hard disk file faxcover.exe> ​​Fax Cover, Fax Cover, FaxQueue.exe> ​​Display Fax Queue Show Fax Quele Faxsend.exe> ​​Fax Wizard for sending faxes Send Fax Wizard FaxSvc.exe> ​​Starts Fax Server boot fax Services fc.exe> ​​Compares Two files or sets of files and their Differences Compare two files Different Find.exe> ​​Searches for a Text String in File or Files Find files Findstr.exe> ​​Searches for Strings In Files Find Row in the file Finger.exe> ​​Fingers a User and Displays Statistics On That User Finger A user and displays a statistical result fixmapi.exe> ​​fix mappi files to fix Mapi file flattemp.exe> ​​enable OR Disable TemporalLector Allow or disable Temporary Directory FontView.exe> ​​Display Fonts in A Font File Displays font forcedos.exe> ​​forces a file to start in dos mode. Forced files running Freecell.exe in DOS mode> Popular Windows Game FreeCell ftp.exe> ​​File Transfer Protocol used to transfer files over a network connection FTP is the gdi.exe> ​​graphic Device interface GUI driven grovel.exe> ​​grpconv.exe> ​​program Manager group Convertor conversion program administrators group help.exe> ​​displays help for Windows 2000 commands to display help hostname.exe> ​​display hostname for machine. display machine hostname ie4uinit.exe> ​​IE5 user install tool IE5 user installation tool ieshwiz.exe> ​​customize folder wizard custom folder Wizard iexpress . .exe>

Create and setup packages for install through installed packages iisreset.exe> ​​Restart IIS Admin Service to restart the IIS service internat.exe> ​​Keyboard Language Indicator Applet keyboard language indicator ipconfig.exe> ​​Windows 2000 IP configuration. To view the IP configuration ipsecmon.exe> ​​IP security monitor IP security monitor ipxroute.exe> ​​IPX routing and source routing control program IPX routing and source routing control program irftp.exe> ​​Setup FTP for wireless communication wireless connection ismserv.exe> ​​Intersite messaging Service to install or remove the Service control Manager Serving JDBGMGR.EXE> Microsoft Debugger for Java 4 Java4 debugger JetConv.exe> ​​Convert a Jet Engine Database Convert Jet Engine Database JetPack.exe> ​​Compact Jet Database. Compressed Jet Database JView.exe> ​​Command-Line Loader for Java Java Command line loader KRNL386.EXE> Core component for Windows 2000 2000 Label.exe> ​​Change Label For Drives Change the Board LCWIZ.EXE> License Compliance Wizard for Local or Remote Systems. License Conforms the Wizard LDIFDE.exe > LDIF CMD LINE Manager LDIF Directory Exchange Command Line Management Licmgr.exe> ​​Terminal Server License Manager Terminal Services License Agreement Management Lights.exe> ​​Display Connection Status Lights Display connection LSMGR.EXE> Windows 2000 License Manager 2000 License Agreement Management Llssrv.exe> ​​Start The License Server Start License Protocol Server LNKSTUB.EXE> Locator.exe> ​​RPC Locator Remote Location Lodctr.exe> ​​LOAD Perfmon Counters Call Performance Count Logoff.exe> Log Current User Off. Log User LPQ.EXE> Displays Status of A Remote LPD Queue Displays the status of the remote LPD print queue, displaying the DIX-based server-based print task lpr.exe> ​​Send a print job to a network Red Direction Print Tasks to printers in the network. Usually used for UNIX client printers to send print tasks to printer servers connected to the NT of the print device.

lsass.exe> ​​LSA Executable and DLL lserver.exe Server DLL run LSA and the Server> Specifies the new DNS domain for the default server to specify the default new DNS domain macfile.exe Server> Used for managing MACFILES management MACFILES magnify.exe> ​​Used to magnify the current screen magnifier makecab.exe> ​​MS Cabinet Maker make CAB file mdm.exe> ​​machine debug Manager machine debug Manager mem.exe> ​​display current memory stats display memory status migpwd.exe> ​​migrate passwords. migrate passwords mmc.exe> Microsoft Management Console Console Mnmsrvc.exe> ​​Netmeeting Remote Desktop Sharing NetMeeting Remote Desktop Sharing Mobsync.exe> ​​Manage Synchronization. Several common RAID levels and comparisons

RAID is the abbreviation of Redundant Array OxpenSive Disk. It is an inexpensive redundant disk array that is a technical implementation of disk arrays in technology, and its purpose is to reduce errors and improve the performance and reliability of storage systems. Common levels have 0, 1, 3, and 5.

1. What is RAID LEVEL 0? RAID Level 0 is the implementation of Data Striping technology, which makes all hard drives a disk array, which can read and write multiple hard drives at the same time, but do not have backup and fault tolerance. It is cheap, and the hard disk is used optimal, but reliability is the worst. With a RAID Level 0 disk array composed of two hard drives as an example, it writes the first and 2nd of the data into the first hard disk, the third and fourth digits are written to the second hard disk ... with this class, So "data segmentation", because the writing action of each disk is done simultaneously, so its storage speed can be several times faster than a single hard drive. However, in this way, there is a hard disk on the disk array. Since it is disconnected to different hard drives, it is broken, and if there is an integrity of the data, if there is no backup of the entire disk array For tape, all data cannot be kept. Therefore, although its efficiency is high, rarely rarely braving the danger of data loss. 2. What is RAID LEVEL 1? RAID Level 1 uses Disk mirror (disk mapping) technology, which is copied to another hard disk for a hard disk, so it has backed up and fault tolerance. Not high, but high reliability. 3. What is RAID LEVEL 3? RAID Level 3 With Byte-Interleaving technology, the hard disk is simultaneously moved at the SCSI control card, and the data used for parity is stored in a particular drive, it has it. Fault-tolerant ability, the use efficiency of the hard disk is installed, which is reduced, and its reliability is better. 4. What is RAID LEVEL 5? RAID Level 5 uses Disk striping (hard disk split) technology, unlike Level 3, it is stored in various hard drives, and each hard disk is controlled by the SCSI control card. Lower operation, fault tolerance, like Level 3, its use efficiency is also installed and then minus one.

The advantages and disadvantages of the RAID level: RAID 0 Access speed is the fastest way to be misleasted RAID 1 full fault tolerance cost high RAID 3 write performance is not multi-task feature RAID 4 has multi-task and fault tolerance Parity disk drive to cause performance bottleneck RAID 5 Multi-task and fault-tolerant feature have Overhead RAID 0 1 / RAID 10 speed fast, completely fault-tolerant cost

Introduction to RAID 5EE Array Level

Introduction to RAID 5EE Array Level

Document content:

In IBM's new array card products, support for a new RAID level ---- RAID 5EE. In the past, we may listen more about RAID 0, RAID 1, RAID 5, and RAID 5E, below, we will introduce the difference between RAID 5EE and several other common RAID levels. RAID 0: Combine multiple disks into a large disk, does not have redundant features, parallel I / O, the fastest speed. It is columns into multiple disks, making a big hard disk. When storing data, it performs the number of data by the number of disks, and then writes the data into these disks. So, in all levels, RAID 0 speed is the fastest. But RAID 0 has no redundancy, if a disk (physical) is corrupted, all data cannot be used. RAID 1: The same disk system is mirrored by the same disk system in two groups, but the speed is not improved, but a single disk is incorrect, the reliability is the highest. RAID 1 is mirror. Its principle is to write the same data on the mirror hard disk while storing data on the main hard disk. When the primary hard disk (physics) is corrupted, the mirror hard disk replaces the work of the main hard disk. Because there is a mirror hard disk to do data backup, the data security of RAID 1 is best in all RAID levels. However, the utilization of its disk is only 50%, which is a level of the lowest disk utilization rate on all RAIDs. RAID 3 stores the principle of data and RAID 0, and RAID 1 is different. RAID 3 is a parity bit that stores data with a hard drive, and data is stored in the rest of the hard disk. It stores the number as in parallelism like RAID 0, but the speed is not RAID 0. If the data disk (physics) is damaged, as long as the bad hard disk is replaced, the RAID control system will rebuild data on the bad disc in the new disk according to the data check digit of the checkboard. With a separate verification disk to protect the data, although there is no mirroring, the hard disk utilization has been greatly improved, which is N-1. However, the disadvantage is that the hard disk of the storage of the calibration is large, because each write operation, the generated check information is written to the disk, and the load of other disks is relatively small, which will have certain performance Impact. RAID 5: RAID 5 is some improvements on the RAID 3, and when the disk write data is written to the array, parity data is evenly placed on each disk in the array, allowing a single disk to be erroneous. RAID 5 is also secure data of the data check digits, but it is not stored in a separate hard drive, but stores the check digits of the data segment to each hard disk. In this way, any hard disk is corrupted, and the damaged data can be rebuilt according to the check digits on the other hard drives. The utilization rate of the hard disk is also N-1, requiring at least 3 hard drives to be implemented. RAID 5E (RAID 5 enhencement): RAID 5E is improved on the RAID 5 level, similar to RAID 5, and the verification information of the data is evenly distributed on the hard drives, but part of each hard disk is not used. The space, this part of the space is not striped, allowing up to two physical hard drives to malfunction. It seems that RAID 5E and RAID 5 add a hot spare disk as if it is similar, in fact, because RAID 5E is distributed on all hard drives, performance will be better than RAID5 plus a hot spare plate. When a hard disk fails, the data on the faulty hard disk is compressed into the unused space on the other hard drive, and the logical disk keeps the RAID 5 level.

RAID 5EE: Compared to RAID 5E, RAID 5EE's data distribution is more efficient, part of each hard disk is used as a distributed hot spare disk, which is part of the array, when a physical hard disk in the array fails, data The reconstruction speed will be faster. In the figure, D represents the data, P represents the check information, and s is the space of the hot standby, which can be seen that the preserved hot standby space is evenly distributed on the hard disk. Use index and statistical features to improve database query performance

Author: TechRepublic.comFriday, March 5 2004 9:51 AM

Special note: In Microsoft's SQL server system, proper optimization of database query functions requires some basic query indexes and performance statistics. How to be familiar with the optimization of the system will achieve the correctness of decision-making. As you have gradually familiar with Microsoft's SQL server database, performance optimization needs will also be further increased. The first step in building a database environment that truly implementing the optimal query function is to know how the SQL server system is working. Although the index is for a particular query, inquiring planning and performance optimization may only require a small amount of cost estimation and comparison, and there is no cost estimate and comparison, but most queries will benefit from implementation of fully optimized. One of the most effective ways to improve query performance is to create a high efficiency index. A good index of a framework avoids the situation of scanning the entire data table when performing query. When creating an index, the SQL server system will automatically measure and store statistics corresponding to the distribution status value associated with the index column. These statistics are often used by the optimizer to assess whether the inquiry strategy is reasonable. There are two types of index: Clustered index and non-clustered indexes, depending on the data set, each type of index has its own unique advantages. The Clustered Index requires the data in the data table to store in order. Because the data has been sorted, the Clustered index is very effective for finding a certain range of index values. This type of index performance is also superior to other types of indexes for row information with unique index values. The Non-Clustered index and the index in the textbook are very similar, and the index is in one position and its data value is in another position. For a query search for a data value, first search the Non-Clustered index, find the location of the data value in the data table, and then obtain data directly from this location. Non-Clustered Index is very useful for exact matching queries. As a commonly used rule, the number of indexes should be as small as possible to reduce the processing process associated with each query. If you want to analyze and optimize the performance of the query, you should measure and collect statistical features of the data. The SQL server system can maintain the data statistical characteristics of the index value. If appropriate, the statistical amount can be performed for non-index values. For performance optimization, database administrators should know how basic statistical concepts, which are defined as follows: Bench: Metrics can exist how many unique values ​​in the data set. Density: The number of metrics only in the data set. The density is obtained by the following method: the number of rows of a given key value divides the total number of data tables. The optimizer will ignore high density indexes. Select rate: Metrics will return the number of query results for a specific query. The selectivity is obtained by the following method: query the number of rows obtained by querying the query. To calculate the relative cost of the query plan, the optimizer requires a valid selection rate to measure. With the changes in data in the column, index and column statistics become no use, which will cause optimal performance without optimal performance when deciding how to handle queries. Therefore, the SQL server system cycle automatically updates these statistics depending on the change in data in the data table. By sampling of these data, the automatic update of this statistics will make the cost to the lowest and do not need to analyze all data. Best performance is designed in a complex database table and specifying an index is a very tricky task. Fortunately, the SQL server system has a built-in adjustment wizard to help you build the best statistics and index collection. To improve the database's query performance, you can provide a script-based suggested list by run the wizard. How to work for the SQL server query optimizer, the more you know, you will know why the specific situation can only be implemented with the proposal to use the wizard. However, for dynamic systems, the best database performance analysis section will need to be updated periodically. Understanding the true meaning of each statistical measure in query index performance will help you have a good knowledge foundation in management decisions.

The index to improve the performance of SQL Server: TechRepublic.comFriday, March 5 2004 9:45 AM

Special note: In Microsoft's SQL Server system, you can improve the query performance of the database through a valid usage index, but the performance improvement depends on the implementation of the database. In this article, you will tell you how to implement indexes and effectively improve the performance of the database. Using indexes in relational databases can improve database performance, this is very obvious. The more indexes are used, the faster the speed from the database system. However, it should be noted that the more indexes are used, the more time it takes to insert new data into the database system. In this article, you will learn about the various types of index supported by Microsoft's SQL Server database, where you will learn how to use different methods to implement indexing, through these different implementation methods, you read in the database Performance is much more than the overall performance of the database. The index defined index is a database of the database. By using an index, when you get data in the database, you can need to scan all the data records in the database, which can improve the performance of the system to obtain data. Using an index can change the organization of data, so that all data is organized in a similar structure so that data can be easily achieved access. The index is created according to the column, which can help the database find the appropriate data according to the value in the index column. Indexed Type Microsoft's SQL Server supports two types of indexes: Clustered indexes and nonclustered indexes. The Clustered index stores data in a physical order in the data table. Because there is only one physical order in the table, there can only be a Clustered index in each table. The Clustered index is a very effective index when looking for data within a range, because these data has been sequenced in the physical order when they are stored. The NonClustered index does not affect the following physical storage, but it is made of a data row pointer. If there is already a Clustered index, the index pointer in Nonclustered will contain the location reference of the Clustered index. These indexes are more prioritized by data, and the scanning speed of these indexes is much better than the actual data sheet. How to implement an index database automatically creates certain indexes. For example, Microsoft's SQL Server system enforces the unique constraint by automatically creating a unique index, which ensures that duplicate data is not inserted in the database. You can also create additional indexes or through SQL Server Enterprise Manager, and SQL Server Enterprise Manager has an index creation template to guide you how to create an index. Get better performance Although the index can bring performance advantages, it will also bring a certain price. Although the SQL Server system allows you to create up to 256 NonClustered indexes in each data table, it is recommended not to use so many indexes. Because indexes need to use more storage spaces on the memory and physical disk drive. In the process of performing the insert declaration, the system performance may be reduced to a certain extent, because in insertion data, it is necessary to insert according to the order of the index, not the data directly in the first available location, so that The more indexes are required to cause the time required to insert or update the declaration. When you create an index using the SQL Server system, it is recommended to refer to the following creation criteria: Correct selection data type: Use some data types in the index to improve the efficiency of the database system, such as int, bigint, smallint, and tinyint These data types are ideal for use in an index because they all take advantage of the same size space and can easily implement comparison operations.

Other data types such as CHAR and VARCHAR are very low, because these data types are not suitable for performing mathematical operations, and the time to perform comparison operations is longer than the above data type. Make sure that the index value is used in the process of use: When the query operation is executed, the column that may be used is only part of the Clustered, and it is especially noted that these data is to be used. These functions may fail when using these data columns as a parameter call function. For example, using the date value as an index, in order to implement the comparison operation, it may be necessary to convert this date value to a string so that this date index value cannot be used during the query. When creating a multi-column index, you need to pay attention to the order of the column: The database will be arranged according to the value of the first column index, and then further sorted according to the value of the second column, sequentially in sequence until the last index sort is completed. Which column is small, which column should be the first index, which ensures that data can be joined for further cross-sorting. The number of columns in the Clustered index: The more columns used in the clustered index, the more the clustered index reference position included in the nonclustered index, the more data that needs to be stored. This will increase the size of the data table containing the index, and will increase the index-based search time. Avoid frequent updating Clustered index data columns: Since the NonClustered index depends on the Clustered index, if the data column constituting the Clustered index is frequently updated, the row locator will result in the row locator stored in NonClustered updates. For all queries associated with these columns, this will result in an increase in performance costs if the record is locked. Separate operation (if possible): For a table, if you need to perform frequent execution insert, update operation, there is a large number of read operations, attempting to separate this table in the case of possible. All inserts and updates can be operated in a table without an index, then copy it to another table, and there is a large number of indexes in this table to optimize the ability to read data. Appropriate reconstruction index: Nonclustered index contains a pointer to the Clustered index, so that the nonclustered index will depend on the Clustered index. When rebuilding the Clustered index, first is to discard the original index, then use Create Index to create an index, or use the CREATE INDEX declaration to declare a part of the index while using the CREATE INDEX declaration. Summary of discarding and creating a few steps will result in multiple rebuild Nonclustered indexes, rather than using the Drop_existing clause, only rebuild a nonclustered index. Wise use of padding factors: Data stores within those continuous memory pages with fixed size. With the addition of new record lines, the data memory page will gradually fill, the system must perform splitting work of the data page, transfer some data to the next new page through this split job. After such splitting, the burden of the system will be increased, and the stored data is broken. The fill factor can maintain the gap between data, which is generally in the creation of an index, and the fill factor of the index has been set. In this way, the number of pages splitages caused by the inserted data can be reduced. Because only the size of the space is maintained when the index is created, the size of the space is not maintained when adding data or updating data. Therefore, it is necessary to reconstruct the index with the filling factor to be sufficiently utilized. The gap caused by the fill factor will result in a decrease in readability, because with the expansion of the database, more and more disk access needs to read data.