error recover deleted messages from Restore Area in GWDR (Reload)

Hello GW faimily!, we are having the following issue when we are trying to recover deleted messages from RELOAD backups.

We are not being able to restore messages from the GAC or also from the GWclient, always with the same reported error in the POA logs.

GrouWise servers version: 24.2.0-145238
GWclient version: 24.2 build 15/03/2024
GWDR Version (Reload): 18.1.0 Build (1810255) - Licensed Software Version: 5.00



GWclient error: [8219] Failed to connect to a remote network resource
Errors in the POA logs:
11:34:05 A3DF Unable to access backup location: /mnt/restore/p2/connect/restore User[ baynet-test]
11:53:08 A3DF User baynet-test tries to restore from backup location: /mnt/restore/p2/connect/restore
11:53:08 A3DF The database function 14 reported error [8201] on ngwguard.db

These are the same error we got if we try to restore the messages for any user from the GAC:
11:54:58 A31E Processing Admin Message (a64f2e92.001)
11:54:58 A31E Processing admin task: Full User Restore
11:54:58 A31E The database function 14 reported error [8201] on ngwguard.db
11:54:59 A31E The database function 14 reported error [8201] on ngwguard.db

But I continue having the same error, no matter where you try to do the restore, everything that implies that the POA must access the RestoreArea ends in the same error



RELOAD nfs configuration:
reload:~ # cat /etc/exports
#/bkp 192.168.135.22(rw,no_root_squash,sync,no_subtree_check) 192.168.135.2w,no_root_squash,sync,no_subtree_check) 192.168.135.24(rw,no_root_squash,sync_subtree_check) 192.168.135.25(rw,no_root_squash,sync,no_subtree_check) 192.1135.26(rw,no_root_squash,sync,no_subtree_check) 192.168.135.27(rw,no_root_squ,sync,no_subtree_check)
/bkp *(rw,no_root_squash,sync)
reload:~ #

At this moment I am working very well in the Opentext team after opening the ticket 02880835 . In a few minutes I will be sharing with you the results of the tests of a very strong hypothesis.
I'll look forward to hearing back from you. THANKS!
Hugs.,

  • Suggested Answer

    0  

    Below I share with you, in case it is useful, the result of additional tests of what was initially a hypothesis, which ended up being the cause of the problem for the moment:

    The Senate of Chile has been using GroupWise for approximately 15 years because it is a rock solid and also a very secure platform. We followed the good practices suggested by Novell of configuring the agents to start as non-root, 

    Ref used: https://www.novell.com/documentation/groupwise18/gw18_guide_install/data/inst_create_agents_linux.html

    HYPOTHESIS:

    Our hypothesis was that perhaps the problem is that all the POAs are running as non-root, this could be the reason why it was not possible to restore messages as long as the attempts were from restore orders launched from the GAC or requested from from a GW client.

    WORKAROUND for confirm the hipothesis:

    1. I've modified the /etc/opt/novell/groupwise/agents/uid.conf . Within this file we had specified the user "hacluster".

    2. After commenting it, I reneamed the uid.run file and raisedUP (start) the POA again so that it could be done as root

    3. After I configured the PAO to start as root, the access to the RestoreArea is right :-)

    Many customer for security reasons surely they are running GroupWise agents in safe mode by launching their agents with non-root, I think it would be important to be able to escalate internally with the Reload team to see what they suggest.

    Another think that I could see is that Reload, although it correctly takes the backups, does not maintain the owners of the DBs. I mention this, because our GroupWise server does this owners "hacluster.root":

    While Reload have this ownerships "root.root":

    In resume, the problem of not being able to recover messages is that the POAs are reunning in safe-mode with non-root, that's why they can't access the Restore Areas. As temporal workaround with the customer needed to execute the the steps detailed above.

    I hope all this information is helpful! Have a great Friday and Weekend!.

    Hugs from ARG.

  • 0   in reply to   

    Good morning GW family how are you?. The team support (ticket 02880835) just confirmed me that what is reported that Reload is now functioning as intended, with the note that when a user is set as non-root, it can lead to certain issues like this. 


    They encourage me to visit the ideas portal and submit suggestions for potential enhancements.

     how are you?, apologies for bother you. Would it be possible to analyze this suggestion allowing GroupWise customer using non-root to recover messages from Restore Areas hosted in RELOAD (GWDR)??.

    Have a great start into this new week!.

    Best regards.,

  • 0   in reply to   

    HI Walter - I am doing well and you are never a bother.  Sadly I am the one that gave the technician you have been working with the information that this is working as designed.  Right now we have only limited resources to work on bug fixes for Reload so without the post in the Ideas community and necessary votes to bring it to the attention of the product manager, this one won't be looked at.  Please go post this in the Ideas community and get others to vote on it.   - Pam

  • 0   in reply to   

    THANKS for your rapid response Pam!!! 

    It is more than understandable, I am very grateful for your sincere response.
    Hugs!!