Hello, i have Zabbix 5.0.23 version using Server and Proxies. Also im using two isolated hosts for the Frontend (httpd 2.4 and php-fpm 7.2).
MySQL version 8.0.15
Since a couple of weeks im getting some strange behavior, PHP pool is dynamic and is starting to grow the amount of php processes. For what I saw, happens when frontend try to check for Problems in problem page or from Problem widget. What happens from Problems page is that sometimes users get 502 http error or from widget Problem the loading weal keeps spinning
If i check the DB current processes im seeing queries that could be running for more than 4000 sec, same query for what i can see.
Like:
SELECT COUNT(DISTINCT a.alertid) AS rowscount,a.eventid FROM alerts a WHERE EXISTS (SELECT NULL FROM events e,functions f,items i,hosts_groups hgg JOIN rights r ON r.id=hgg.groupid AND r.groupid=22 WHERE a.eventid=e.eventid AND e.objectid=f.triggerid AND f.itemid=i.itemid AND i.hostid=hgg.hostid GROUP BY f.triggerid HAVING MIN(r.permission)>0 AND MAX(r.permission)>='2') AND EXISTS (SELECT NULL FROM actions aa WHERE a.actionid=aa.actionid AND aa.eventsource='0') AND (a.userid IS NULL OR EXISTS (SELECT NULL FROM users_groups ug WHERE ug.userid=a.userid AND ug.usrgrpid=22)) AND a.eventid IN (171549,194437,194454,194817,208536,208725,209682, 209705,209724,210026,213358,214582,214596,214872,2 17713,217740,217887,217904,218302,219519,234691,23 5012,235013,235064,235077,235089,235102,235114,571 682,571904,574575,574576,574865,574882,574908,5749 20,769019,1225248,1226534,1226554,1226578,1226595, 1227052,1227053,1227420,1392400,1392412,1392433,13 93034,1395466,1395484,1395504,1395522,1395596,1395 597,1744261,1744265,1745075,1745088,1745105,174512 3,1745158,1745160,1766651,1766673,1766691,1766708, 1766748,1766749,5185522,5185992,5188826,5188973,51 89014,5189118,5189317,5189318,5335130,5335132,5335 146,5335149,5335153,5335155,6332586,6332852,633296 7,6418691,6962830,6963514,6963754,6965597,6965666, 6965705,6965751,6966198,6966200,6971792,7795968,77 96210,7796698,7828035,7828036,7828743,7828828,7828 854,7828919,7932716,10228697,10228699,10228701,112 56798,11256805,11263503,11263505,11264675,11264711 ,11264747,11264775,11264801,11264873,12445280,1276 3819,14473153,14569966,14578805,14579527,16130617, 16130622,16132529,16132575,16132610,16132643,16132 685,16132716,16132759,16132787,16133016,16133058,1 6133092,16133130,16133165,16133196,16133233,161332 71,16136920,16142526,16142565,16142599,16142638,16 142672,16142706,17587540,17589846,17590024,1759020 4,17590363,18506404,18507662,18975552,19595605,200 93620,20200813,21114017,21807745,21943848,21971016 ,21976022,21976024,21976028,21976291,21977800,2205 5301,22136008,22238711,22238712,22338318,22352482, 22360609,22381539,22422079,22436726,22439318,22441 113,22449664,22452164,22453390,22456240,22457243,2 2461914,22461947,22461994,22462008,22462177,224630 04,22463568,22463813,22463861,22463919,22464139,22 464190) GROUP BY a.eventid
These queries come from the Frontend hosts, the workaround im doing is restarting httpd and php services, that will clear the queries but eventually the same situation will happen again.
Nothing has been changed, maybe more users are using the tool but not more than that. Strange that only happens with users with User or Admin privileges, SuperAdmin looks good.
Actually this problem started to happen some weeks ago with version 5.0.7, investigating i found a thread (similar to my issue) that was fixed in v 5.0.15 https://support.zabbix.com/browse/ZBX-16822
For that reason ive upgraded into latest 5.0.23 but issue is still there.
Do you know what could be happening? attaching a screenshot from MySQL workbench, you can see there that is same query running from different PHP connections.
Thanks
MySQL version 8.0.15
Since a couple of weeks im getting some strange behavior, PHP pool is dynamic and is starting to grow the amount of php processes. For what I saw, happens when frontend try to check for Problems in problem page or from Problem widget. What happens from Problems page is that sometimes users get 502 http error or from widget Problem the loading weal keeps spinning
If i check the DB current processes im seeing queries that could be running for more than 4000 sec, same query for what i can see.
Like:
SELECT COUNT(DISTINCT a.alertid) AS rowscount,a.eventid FROM alerts a WHERE EXISTS (SELECT NULL FROM events e,functions f,items i,hosts_groups hgg JOIN rights r ON r.id=hgg.groupid AND r.groupid=22 WHERE a.eventid=e.eventid AND e.objectid=f.triggerid AND f.itemid=i.itemid AND i.hostid=hgg.hostid GROUP BY f.triggerid HAVING MIN(r.permission)>0 AND MAX(r.permission)>='2') AND EXISTS (SELECT NULL FROM actions aa WHERE a.actionid=aa.actionid AND aa.eventsource='0') AND (a.userid IS NULL OR EXISTS (SELECT NULL FROM users_groups ug WHERE ug.userid=a.userid AND ug.usrgrpid=22)) AND a.eventid IN (171549,194437,194454,194817,208536,208725,209682, 209705,209724,210026,213358,214582,214596,214872,2 17713,217740,217887,217904,218302,219519,234691,23 5012,235013,235064,235077,235089,235102,235114,571 682,571904,574575,574576,574865,574882,574908,5749 20,769019,1225248,1226534,1226554,1226578,1226595, 1227052,1227053,1227420,1392400,1392412,1392433,13 93034,1395466,1395484,1395504,1395522,1395596,1395 597,1744261,1744265,1745075,1745088,1745105,174512 3,1745158,1745160,1766651,1766673,1766691,1766708, 1766748,1766749,5185522,5185992,5188826,5188973,51 89014,5189118,5189317,5189318,5335130,5335132,5335 146,5335149,5335153,5335155,6332586,6332852,633296 7,6418691,6962830,6963514,6963754,6965597,6965666, 6965705,6965751,6966198,6966200,6971792,7795968,77 96210,7796698,7828035,7828036,7828743,7828828,7828 854,7828919,7932716,10228697,10228699,10228701,112 56798,11256805,11263503,11263505,11264675,11264711 ,11264747,11264775,11264801,11264873,12445280,1276 3819,14473153,14569966,14578805,14579527,16130617, 16130622,16132529,16132575,16132610,16132643,16132 685,16132716,16132759,16132787,16133016,16133058,1 6133092,16133130,16133165,16133196,16133233,161332 71,16136920,16142526,16142565,16142599,16142638,16 142672,16142706,17587540,17589846,17590024,1759020 4,17590363,18506404,18507662,18975552,19595605,200 93620,20200813,21114017,21807745,21943848,21971016 ,21976022,21976024,21976028,21976291,21977800,2205 5301,22136008,22238711,22238712,22338318,22352482, 22360609,22381539,22422079,22436726,22439318,22441 113,22449664,22452164,22453390,22456240,22457243,2 2461914,22461947,22461994,22462008,22462177,224630 04,22463568,22463813,22463861,22463919,22464139,22 464190) GROUP BY a.eventid
These queries come from the Frontend hosts, the workaround im doing is restarting httpd and php services, that will clear the queries but eventually the same situation will happen again.
Nothing has been changed, maybe more users are using the tool but not more than that. Strange that only happens with users with User or Admin privileges, SuperAdmin looks good.
Actually this problem started to happen some weeks ago with version 5.0.7, investigating i found a thread (similar to my issue) that was fixed in v 5.0.15 https://support.zabbix.com/browse/ZBX-16822
For that reason ive upgraded into latest 5.0.23 but issue is still there.
Do you know what could be happening? attaching a screenshot from MySQL workbench, you can see there that is same query running from different PHP connections.
Thanks
Comment