When you visit a page with syslog on it, it seems to peg the mysql server because you have too many syslog entries for that device.
Step one: ./validate.php
Step two: make sure you are running ./daily.sh in your cron job to clean up old syslog entries.
Step three: if you don’t care about history and don’t want to wait for daily.sh to cleanup, you can truncate the entire syslog table to fix the issue for now.
I tried to run daily.sh, but it seems to hang when cleaning up the DB. After a while mysql crashed and I had a really hard time restarting the VM. It took 15min for Vcenter to reset it.
uma@libre-nms:/opt/librenms$ sudo ./validate.php
PHP Notice: ob_end_clean(): failed to delete buffer. No buffer to delete in /opt/librenms/validate.php on line 163
Ok, the table has a large number of rows
SELECT COUNT() FROM syslog;
±----------+
| COUNT() |
±----------+
| 168159080 |
±----------+
1 row in set (9 min 29.95 sec)
I truncated the whole table and everything looks good now, Thanks for the help!
The problem was that our firewalls were sending a lot of syslog messages to Libre, we have changed that now and the syslog table does not grow as much, Earlier it grew with 1000’s of entries every minute