GridLink Farm Recovery

When a server has stopped working due to an Operating System or hardware failure, and it needs to be rebuilt or removed from the farm, the farm may not work any more and you will have to recover it from the server failure. In this article, we will provide step-by-step farm recovery instructions in various scenarios.

1. Understand Farm Profile

A Farm Profile is a collection of configurations including the list of server computers on the farm and settings for the GridLink controller service running on the server.

2. Understand Farm Profile Host

Every farm has a server called Profile Host, which stores a number of Farm Profiles. e.g. If you have a farm with 2 servers (MACHINE1 and MACHINE2) and MACHINE1 is the Profile Host. You may have created the following profiles:

Profile1: "MACHINE1 Only"

Profile2: "MACHINE2 Only"

Profile3: "MACHINE1 and MACHINE2 - For All Users"

Profile4: "MACHINE1 and MACHINE2 - For Administrator Only"

MACHINE1 (Profile Host) stores all the 4 profiles, plus the profile that is currently in use.

MACHINE2 (non-Profile Host) only stores the profile that is currently in use.

Using GridLink Utility, you can switch from one profile to another for different purposes. Every time when you select a new profile, all servers included in this profile will receive the updated settings. Each server has the farm settings specified in the current farm profile. Only the Profile Host has all the profiles. They are saved in database file GCDB.AXS. Because a Profile Host is usually a master server, so this file also contains all records of remote jobs that have been run on this server.

3. Back up Farm Profiles

It is very important to back up your Farm Profiles (in another word, the file GCDB.AXS on the Profile Host) so that you can completely recover the farm in case of a server failure.

You can use the GridLink built-in Active Backup feature to back up your Farm Profiles. For more information on using Active Backup, please refer to topic "Active Backup".

4. Farm Recovery

NOTE: We assume that there's no AXIS job running on the farm during this recovery process.

4.1 Non-Profile Host Failure

If the failed server is not the Profile Host, then the recovery is very easy.

4.1.1 The server is not removed

The problem has been fixed, but the OS has been reinstalled. In this case, you need to reinstall GridLink service on this server.

Run GridLink Utility (please note that the user account under which the utility is run needs to have the local administrator's right), connect to the Profile Host to retrieve the Farm Profile, and you will see that the utility failed to connect to this server. Right-click on this server, and select the menu "Service - Install Service" and then follow the instructions.

4.1.2 The server has been removed

In this case, you need to remove this server from the Farm Profile.

Run GridLink Utility, connect to the Profile Host to retrieve the Farm Profile, and you will see that the utility failed to connect to this server. Right-click on this server, and select the menu "Remove server from the farm" to remove it from the Farm Profile.

4.2 Profile Host Failure

4.2.1 The server is not removed

The problem has been fixed, but the OS has been reinstalled. In this case, you need to reinstall GridLink service on this server.

Step 1. Run GridLink Utility (please note that the user account under which the utility is run needs to have the local administrator's right).

Step 2. In the log on screen, input the original Profile Host name in the "Profile Host Name" edit box.

Step 3. Click on the "Service" button, select "Install the service" and then click on "OK" button and follow the instructions.

Step 4. If you have GCDB.axs backed up, then do the following:

1) Click on the "Service" button, select "Stop the service" and then click on "OK".

2) After the service has been stopped, copy GCDB.axs to the GridLink installation directory on the server to overwrite the existing file.

3) Come back to the utility, click on the "Service" button, select "Start the service" and then click on "OK".

4) After the service has been started, input the password and click on "OK" button in the log on screen.

5) Select one farm profile if you have multiple profiles.

6) Click on "OK" button.

7) You will see the original server list and the settings in the farm.

Now, the farm has been recovered completely.

Step 5. If you don't have GCDB.axs backed up, then do the following:

1) In the log on screen, input the password and click on "OK" button, and you will see "Create Farm Profile" dialog.

2) Input the farm name.

3) Select "This new farm uses configurations retrieved from an existing server" and specify the name of another existing server in the farm. This server has to have GridLink service running. In the above example, this server is MACHINE2.

4) Make sure "Only retrieve the farm settings, excluding the list of computers" is unchecked.

5) Click on "OK" button.

6) You will see the server list and the settings retrieved from MACHINE2.

Now, the farm has been recovered.  (Please note that if you have multiple farm profiles, only the one that was being used on MACHINE2 can be retrieved. You will lose all other profiles.)

4.2.2 The server has been removed

The recovery process is more difficult in this case, because you will need to designate a new Profile Host.

Step 1: Log on to MACHINE2, go to the GridLink installation directory and run FirstAid.exe.

Step 2: It should show the original Profile Host name MACHINE1. Clear the Profile Host name in the edit box.

Step 3: Click on "OK", and then click on "Yes" to confirm to restart the service.

Step 4: If you have other servers in the farm, repeat Step 1 - 3 for each of them.

Step 5: Run GridLink Utility.

Step 6: In the log on screen, input the newly designated Profile Host name in the "Profile Host Name" edit box. Let's take MACHINE2 as the example.

Step 7: Input the password and click on "OK" button, and you will see "Create Farm Profile" dialog. (assume that this server was never used as a Profile Host before.)

Step 8: Select "This new farm uses configurations retrieved from an existing server" and specify MACHINE2 itself.

Step 9: Make sure "Only retrieve the farm settings, excluding the list of computers" is unchecked.

Step 10: Click on "OK" button.

Step 11: You will see the server list and the settings retrieved from MACHINE2.

Now, the farm has been recovered. (Please note that if you have multiple farm profiles, only the one that was being used on MACHINE2 can be retrieved. You will lose all other profiles.)

5. Farm Recovery with GGY's Assistance

In the following case:

1) You have created multiple Farm Profiles.

2) You have GCDB.axs backed up from the Profile Host.

3) The Profile Host failed and has been removed.

If you wish to recover all Farm Profiles that you created before, please follow the instructions in 4.2.2 and then send the following 2 files to GGY:

1) GCDB.axs from the original Profile Host

2) GCDB.axs from the new Profile Host

We will manually recover all Farm Profiles and send a new GCDB.axs to you. Upon receiving this file, please continue with the following steps:

Step 1. Run GridLink Utility (please note that the user account under which the utility is run needs to have the local administrator's right).

Step 2. In the log on screen, input the new Profile Host name in the "Profile Host Name" edit box.

Step 3. Click on the "Service" button, select "Stop the service" and then click on "OK".

Step 4: After the service has been stopped, copy GCDB.axs to the GridLink installation directory on the new Profile Host to overwrite the existing file.

Step 5: Come back to the utility, click on the "Service" button in the log on screen, select "Start the service" and then click on "OK".

Step 6: After the service has been started, input the password and click on "OK" button in the Log on screen.

Step 7: Select one farm profile from the list, and click on "OK" button.

Step 8: You will see the server list and the settings in this farm profile.

Now, the farm has been recovered completely.