Hi all,
we have a absolutely fine running zabbix installation, but since about a week, all out of a sudden, history syncer is causing myriads of false positives (.nodata - Server is down) when/during its run. Please see the attached screenshot. History cache is filling up rapidly.
NVPS are ~770, DB is backed on a SSD-RAID1, 55GB in size.
Here some mysqltuner output:
[--] Reads / Writes: 77% / 23%
[--] Total buffers: 60.3G global + 125.2M per thread (350 max threads)
[OK] Maximum possible memory usage: 103.1G (81% of installed RAM)
[OK] Slow queries: 0% (1/390M)
[!!] Highest connection usage: 86% (304/350)
[OK] Key buffer size / total MyISAM indexes: 800.0M/126.0K
[OK] Key buffer hit rate: 100.0% (852K cached / 2 reads)
[!!] Query cache efficiency: 0.0% (0 cached / 165M selects)
[OK] Query cache prunes per day: 0
[OK] Sorts requiring temporary tables: 0% (2 temp sorts / 155M sorts)
[!!] Joins performed without indexes: 6836172
[OK] Temporary tables created on disk: 21% (7M on disk / 34M total)
[OK] Thread cache hit rate: 99% (841 created / 8M connections)
[OK] Table cache hit rate: 75% (419 open / 554 opened)
[OK] Open file limit used: 0% (19/5K)
[OK] Table locks acquired immediately: 100% (422M immediate / 422M locks)
[OK] InnoDB data size / buffer pool: 55.0G/58.6G
Any thoughts what's wrong here? Are the SSD's becoming a bottleneck?
Thanks
Edit: we are on 2.2.13
we have a absolutely fine running zabbix installation, but since about a week, all out of a sudden, history syncer is causing myriads of false positives (.nodata - Server is down) when/during its run. Please see the attached screenshot. History cache is filling up rapidly.
NVPS are ~770, DB is backed on a SSD-RAID1, 55GB in size.
Here some mysqltuner output:
[--] Reads / Writes: 77% / 23%
[--] Total buffers: 60.3G global + 125.2M per thread (350 max threads)
[OK] Maximum possible memory usage: 103.1G (81% of installed RAM)
[OK] Slow queries: 0% (1/390M)
[!!] Highest connection usage: 86% (304/350)
[OK] Key buffer size / total MyISAM indexes: 800.0M/126.0K
[OK] Key buffer hit rate: 100.0% (852K cached / 2 reads)
[!!] Query cache efficiency: 0.0% (0 cached / 165M selects)
[OK] Query cache prunes per day: 0
[OK] Sorts requiring temporary tables: 0% (2 temp sorts / 155M sorts)
[!!] Joins performed without indexes: 6836172
[OK] Temporary tables created on disk: 21% (7M on disk / 34M total)
[OK] Thread cache hit rate: 99% (841 created / 8M connections)
[OK] Table cache hit rate: 75% (419 open / 554 opened)
[OK] Open file limit used: 0% (19/5K)
[OK] Table locks acquired immediately: 100% (422M immediate / 422M locks)
[OK] InnoDB data size / buffer pool: 55.0G/58.6G
Any thoughts what's wrong here? Are the SSD's becoming a bottleneck?
Thanks
Edit: we are on 2.2.13
Attached Files