This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Upgrade ChangeAuditor and connecting to a new database

Hi,

Because of large amount of events that couldn't be written to the database in time I decided to upgrade the system from version 6.6 to version 6.9.4

Since I had a large database and many events waiting to be written I've decided also to detach my current database and creating a new one.

After detaching the database I've started the coordinator upgrade leaving the same installation name as "Default" and specified a new database name (database catalog name) that is different from the old one.

The upgrade finished, created the new database but the coordinator remains in "Initializing" status to long so I've opened the coordinator log file and noticed that the coordinator is Replaying log files from C:\ProgramData\Quest\ChangeAuditor folder (txlog files) many of them are quiet old (2 months ago).

I don't see that my new database files are changing in size and they aren't changed from the moment the coordinator started the replaying process.

My questions:

1) What are those files and what the replaying process are doing with them in the new empty created database ?

2) Should I had to delete all the txlog from this folder prior of the coordinator upgrade ? I didn't see any of that in any of the procedures.

3) Should I leave it to replay all the txlog files before I'll continue with the client upgrade and the agents upgrades ?

4) Any suggestions according to my situation ?

Thank you in advanced for any help and suggestion.

Regards,

Parents
  • Hi Avi,

    Feedback below...

    1) What are those files and what the replaying process are doing with them in the new empty created database ?

    These are files containing pending transactions waiting at the previous coordinator shutdown. Sounds like coordinator was overloaded or existing database not accepting more events leading to backlog of pending transactions for some time now (you mentioned the files were months old).

    2) Should I had to delete all the txlog from this folder prior of the coordinator upgrade ? I didn't see any of that in any of the procedures.

    They will replay eventually given time. Else, if there are many and you do not wish to allow them to replay you could stop coordinator, rename the "c:\programdata\quest\changeauditor" folder to "c:\programdata\quest\changeauditor_old\", drop the old db to ensure consistency, start the coordinator/create new db.

    3) Should I leave it to replay all the txlog files before I'll continue with the client upgrade and the agents upgrades ?

    You can proceed to upgrade the client/agents without issue while these transactions are replaying without issue.

    4) Any suggestions according to my situation ?

    I would suggest to simply let the transaction logs replay depending on the # of them found there. If there are many and we're not concerned with these transactions, it would be faster to bring coordinator online using the feedback in #2 above.

    Thanks,
    Chris

Reply
  • Hi Avi,

    Feedback below...

    1) What are those files and what the replaying process are doing with them in the new empty created database ?

    These are files containing pending transactions waiting at the previous coordinator shutdown. Sounds like coordinator was overloaded or existing database not accepting more events leading to backlog of pending transactions for some time now (you mentioned the files were months old).

    2) Should I had to delete all the txlog from this folder prior of the coordinator upgrade ? I didn't see any of that in any of the procedures.

    They will replay eventually given time. Else, if there are many and you do not wish to allow them to replay you could stop coordinator, rename the "c:\programdata\quest\changeauditor" folder to "c:\programdata\quest\changeauditor_old\", drop the old db to ensure consistency, start the coordinator/create new db.

    3) Should I leave it to replay all the txlog files before I'll continue with the client upgrade and the agents upgrades ?

    You can proceed to upgrade the client/agents without issue while these transactions are replaying without issue.

    4) Any suggestions according to my situation ?

    I would suggest to simply let the transaction logs replay depending on the # of them found there. If there are many and we're not concerned with these transactions, it would be faster to bring coordinator online using the feedback in #2 above.

    Thanks,
    Chris

Children
No Data