Truncating Oracle archived logs issue

[KOPFteam]

As I wrote before: Since I activeted archivelogmode in our Oracle DBs, our Veeam 9 Backupjob is working without problems.
On the Oracle tap in Guest Processing, I choose "Truncate logs over 5 GB". At the end of each run, Veeam gives me a green checkmark "Truncating Oracle archived logs".
Everything looks nice, but I have 5.2 GB archivelogs now, 2.6 GB for each of the two Oracle DBs in that VM. Is that limit (5 GB in my case) an overall limit (then something is wrong) or is it a per database limit?

Thanks,
Florian

[KOPFteam]

5.6 GB archivelogfiles today and the Veeam 9 Backup job again gave me a green checkmark "Truncating Oracle archived logs". Why doesn't it remove the oldest archivelogs to keep the logs under 5 GB?

[foggy]

Florian, this is a per database limit.

[svens]

@Florian
I've a opened case ID# 01691199 since 10 days with the problem that Veeam B&R does not delete backuped archive Log files on one Suse Linux machine with Oraccle 11.
We have other maschines on that the truncate of logfiles is working.

[KOPFteam]

Thanks Foggy, good to know!

I'll change the setting to 3 GB and let you know what happens when it reaches that limit in a few days.

[apronk]

@Florian:
I created a workaround for VBR 8 a while back that you can use if it doesn't work.
Here is an Oracle log cleanup script that I posted here:veeam-backup-replication-f2/feature-req ... ml#p143731
Create a Scheduled Task that triggers on the Oracle.VSSWriter.<SID> event with Event ID 4107, informational only.
This works perfectly, since that specific event has this information: "VSS-04107: successfully backed up component Oracle Database (FULL)"
And then the script runs fully without issue, the script cleans up all archivelogs.

[KOPFteam]

It looks like Veeam is only truncating the archivemodelogs of one of my Oracle instances!

I have 4 instances running on that server. Two are production DBs, those are running in archivelogmode. Two other DBs are for test purposes only and are running in nonarchivelogmode.

I have defined "log_archive_dest" in the databasename.ora for the two databases that are running in archivelogmode. These paths are pointing to two subfolders (one for each DB) of a logfolder.

I have defined a limit of 3 GB in the Veeam 9 backup job.

One of the two archivelogfolders currently has a size of 3,04 GB and the oldest logfiles are gone, so here everything seems to work. The other archivelogfolder currently has a size of 3,44 GB and all logfiles are still there.

What's wrong here?

Thanks,
Florian

[svens]

In my case, there was a missing entry in /etc/oratab. Veeam does only truncate DBs that are in /etc/oratab.

[foggy]

There is also a known issue when Veeam B&R is not able to locate DBs in oratab in case inventory.xml is not valid (contains wrong paths). Could be the culprit in this case as well (will be addressed in the next update).

[KOPFteam]

Nothing has changed. Still the logfiles of one of the Oracle databases are truncated every night and the logfiles of the second Oracle database aren't touched by Veeam. I will install Veeam 9 Update 1 later today and let you know what happens.
BTW: We're running the Oracle databases on Windows not on Linux, so unfortunately I can't test the recommendations some of you gave above.

[veremin]

As Alexander's mentioned, it might be worth checking inventory.xml for the presence of incorrect paths. If everything looks good and Update 1 doesn't make any difference, kindly, open a ticket with our support team. Thanks.

[KOPFteam]

Update 1 didn't help

I had a ticket #01683017, but it was opened for the original problem (nonarchivemode instances were stopped during backup). Because I activated archivemode in the meantime, support closed the ticket. But the backup of my Oracle database is still not perfect.

I leaned a few things today:

• I have 1 (one) Oracle database running. That's why there is just one path in the inventoy.xml and this path is correct.
• I have 4 (four) instances running in this database. 2 of them are used for production and are running in archivelogmode. The other 2 are for testing and are running in nonarchivelogmode.
• For each of the 2 instances running in archivelogmode, there is an initxxx.ora. In both I added a line "log_archive_dest" with a path (D:\ORACLE\Archivelog\BP71 and D:\ORACLE\Archivelog\AM83). This is working, the logs of the instances are created in those folders.
• Veeam 9 is only truncating the log in one of those two folders. It doesn't touch the second one

Any ideas? Should I reopen the ticket with support? Or open a new ticket?

Thansk,
Florian

[foggy]

Yes, please contact support with any technical issues, as a closer look at your setup is required.