Cacti poller, spine errors

[eholz1]

Hello All,
I hate to bring up this topic, but - yes, it is related to poller warnings, error 2013, and thread issues, etc.
We had a power fail, I run cacti 1.2.21 on an Oracle 7 vm. The power fail completely destroyed the cacti
server, more or less, at least in terms of cacti, which worked like a champ. Also, part of the issue might have been, that I applied Oracle updates to the server.

I could not get the server running - EVEN from a Veeam restore! Ouch!
So I updated to cacti-1.2.23, with appropriate changes. But still the cacti app has many poller overruns, thread errors, etc.

Might there be a way to fix this? Without starting all over, that is. cron runs poller every minute, and spine runs every 30 seconds.

I am open to suggestions.
thanks for an amazing product.
eholz1

[cigamit]

The first thing I would check after a power failure is whether any of the db tables are corrupted. There is a script in the cli directory called "repair_database.php" I would run that from the command line and ensure all the tables are good. After that, I would post the errors and the STAT line from the cacti log so we aren't guessing in the dark about what might be wrong.

[TheWitness]

There was a bug introduced in the latest spine that caused a segfault with some older devices, or devices that did not have a valid oid. It's been patched in the 1.2.x branch by a user. If you could update to that and try again, I think the problem will be solved.

[eholz1]

Hello Cacti experts,

Thanks for the reply. I appreciate it. I will look into your suggestions.

Fingers crossed.

Eholz1