Start a Conversation

Unsolved

This post is more than 5 years old

1416

November 1st, 2016 19:00

SyncIQ - Can I modify without having to resync

Community,

Here is the scenario.

  1. A syncIQ policy was created on directory i.e. /ifs/data/synciq to sync data from Source Cluster to Target Cluster.
  2. There are multiple child directories within /ifs/data/synciq
  3. The initial sync for policy /ifs/data/synciq completes its over 250 + TB takes weeks to complete over a 1GB network.
  4. Customer A, and B decide they no longer want backups of their data located at /ifs/data/synciq/FolderA and /ifs/data/synciq/FolderB
  5. An exclusion was created on the SyncIQ policy for /ifs/data/synciq to now exclude folders /ifs/data/synciq/FolderA and /ifs/data/synciq/FolderB from continuing to sync its deltas.


Although I am pretty sure I know the answer to this question.  I have to ask just in case I am not aware of things that can be done but not recommended .  I am ok with disclaimers.


Question is:  What steps can we taken to reclaim the space folders /ifs/data/synciq/FolderA and /ifs/data/synciq/FolderB are taking up on the Target cluster?  Is it even possible to force the target sync location into write mode and remove the excluded folders so that we can reclaim its space?  Is it possible to do all of this and maintain the integrity of the sync policy and the data within without having to resync the ENTIRE directory over again?


***PLEASE NOTE***

  1. We want the best non disruptive to the users solution.
  2. We don't want to have to resync the entire /ifs/data/synciq its too large and would take weeks if not months.

Thank you,

Message was edited by: chjatwork

3 Apprentice

 • 

592 Posts

November 2nd, 2016 10:00

You may be interested in "target aware" sync only via CLI.

https://www.emc.com/collateral/hardware/white-papers/h8224-replication-isilon-synciq-wp.pdf

# isi sync policy modify --policy --diff_sync=on

# isi sync policy ls -v | grep aware

     Target aware initial sync: yes

2 Intern

 • 

356 Posts

November 3rd, 2016 04:00

Phil,

I will pass this information to the teammate handling this issue.  The only thing I am concerned about is that it states that the association between the "policy and target has been broken and is then re-established."  I am guessing that its eluding that we would intentionally brake this association in order to perform this task.  Not having a real life setup to test these steps sort of prevents us from taking such steps, but if we find ourselves in a place where we can effectively perform a test of this magnitude on a much smaller scale it would prove to be interesting.

Any other Ideas by chance that may not be covered in the best practices, but should work?

Thank you,

No Events found!

Top