cancel
Showing results for 
Search instead for 
Did you mean: 

Migrating Data under EV8

Hello,

I was wondering if it's possible to easily migrate archived ReadOnly data in EV 8. In the scenario, the data has been archived to a Net App appliance. This data should be moved to the iTernity iCAS solution.

Does EV 8 make it easily posssible to adjust the paths to the new archive system? Or do the archived objects have dependencies that make such a change difficult?
 

Thank you!

bjorn

1 Solution

Accepted Solutions
Accepted Solution!

This is a difficult one since

This is a difficult one since i know next to nothing about the iTernity iCAS, since most Content Addressable Storage systems require specific migration tools (For instance if you are going to a Centera, there are NTFS to Centera Migration tools)

But from what i can tell, the iCAS can either be written to API (like EV does when talking to a centera) or uses a native File System interface through FSG (File System Gateway)

If it is using the FSG then supposedly it should act as a regular windows interfaceable NTFS or CIFS type of location (i honestly don't know how true this is, considering its all press-release type informational stuff and nothing on the technical side more than that)

IF that is true then it would simply be
1. Backup your Vault store and directory databases
2. Put your EV in to Read Only mode via the Vault Admin Console so no new storage can be written
    (Best would be to stop your entire services)
3. Copy the data from the NetApp across to your iTernity iCAS
4. Update the PartitionEntry table to reflect the new location
5. Restart the Directory and Storage Locations
6. Test that you can retrieve items without issue (if you can , make sure the NetApp is inaccessible to remove any doubt that it might still be accessing the NetApp and not the iCAS)
7. Once you are sure you can open, retrieve and restore etc, take EV out of readonly mode
8. Attemp to archive a couple of items and retrieve them and make sure no errors exist

A technote can be found here:
http://www.symantec.com/business/support/index?page=content&id=TECH35742

If it doesn't work like i think it does or you don't want to use the FSG, then you would need to talk to Symantec, and they will most likely suggest you use TransVault or Move Archive or some other such solution.

So what you would do is create a new Vault Store whose location is on the new server and then start moving the archives across from the netapp to the iCAS, and then once thats completed, remove the netapp and delete the old vault stores

https://www.linkedin.com/in/alex-allen-turl-07370146

View solution in original post

5 Replies
Accepted Solution!

This is a difficult one since

This is a difficult one since i know next to nothing about the iTernity iCAS, since most Content Addressable Storage systems require specific migration tools (For instance if you are going to a Centera, there are NTFS to Centera Migration tools)

But from what i can tell, the iCAS can either be written to API (like EV does when talking to a centera) or uses a native File System interface through FSG (File System Gateway)

If it is using the FSG then supposedly it should act as a regular windows interfaceable NTFS or CIFS type of location (i honestly don't know how true this is, considering its all press-release type informational stuff and nothing on the technical side more than that)

IF that is true then it would simply be
1. Backup your Vault store and directory databases
2. Put your EV in to Read Only mode via the Vault Admin Console so no new storage can be written
    (Best would be to stop your entire services)
3. Copy the data from the NetApp across to your iTernity iCAS
4. Update the PartitionEntry table to reflect the new location
5. Restart the Directory and Storage Locations
6. Test that you can retrieve items without issue (if you can , make sure the NetApp is inaccessible to remove any doubt that it might still be accessing the NetApp and not the iCAS)
7. Once you are sure you can open, retrieve and restore etc, take EV out of readonly mode
8. Attemp to archive a couple of items and retrieve them and make sure no errors exist

A technote can be found here:
http://www.symantec.com/business/support/index?page=content&id=TECH35742

If it doesn't work like i think it does or you don't want to use the FSG, then you would need to talk to Symantec, and they will most likely suggest you use TransVault or Move Archive or some other such solution.

So what you would do is create a new Vault Store whose location is on the new server and then start moving the archives across from the netapp to the iCAS, and then once thats completed, remove the netapp and delete the old vault stores

https://www.linkedin.com/in/alex-allen-turl-07370146

View solution in original post

Thanks for your elaborate

Thanks for your elaborate answer. We will try this today and let you know if it worked.

bjorn

Oh, and one additional

Oh, and one additional remark: Yes, this is using the File System Gateway, which works like NetApp in SnapLock mode or EMC Celera in Worm mode

bjorn

Good answer by JW, just two

Good answer by JW, just two things to add:

  1. You should make sure that there are no items left in the "awaiting backup" status (check http://localhost/enterprisevault/usage.asp on the ev server). Reason: If there are Items left the path of these files is in a table in the database. So after moving the files and editing the DB there will be entries in the DB pointing to the old direction which will not be cleared as EV will not be able to find the files there
  2. One of the easy way to move data without a large downtime is by using robocopy two times: While the site is still in production mirror the content of your partitions to the new storage. When you perform the change you can use robocopy again with the /MIR option as this will only sync the changes since the last robocopy execution. This should only be a relatively small amount of data and thus will be completed in a much shorter time compared to performing the whole copy process during the downtime.

Just saw your last post

Just saw your last post regarding the WORM mode. Then you have to be careful about the second point in my post: To maintain consistency the /MIR switch for robocopy also needs to delete files that were copied in the first run, but are not present any more in the second run. This can happen for various reasons (deletion of elements, collections etc.)

Robocopy will not be able to delete files in the second run when the device is in worm mode - keep that in mind!