i have some problems with the logcache.
Due to problems descripbed here: Issues with Livestatus LMD and Thruk
i used a nightly version to build logcache and then went back to 2.16-2.
Mysql = Server version: 5.6.37-82.2 Percona Server (GPL), Release 82.2
Storage-engine = myisam (due to heavy string search)
On logcacheupdate/import the slowes querys are always on plugin_output table:
- SELECT output_id FROM `a6897_plugin_output` WHERE output = 'Warning: Return code of 255 for check of service \'check_ntp_peer\' on host \'testhost\' was out of bounds.' LIMIT 1;
- # Time: 170920 17:57:18
- # User@Host: thruk[thruk] @ localhost  Id: 194
- # Schema: thruk Last_errno: 1317 Killed: 1053
- # Query_time: 41.417401 Lock_time: 0.000000 Rows_sent: 0 Rows_examined: 0 Rows_affected: 0
- # Bytes_sent: 44
We have around 280k services and 9k hosts... so of course this table is very very big.
Why are plugin outputs stored redundant in table?
I just did a query and most of the output_id's are not inside of the log table.
Is there any need to store things redundant in the same table?