Lots of performance alerts after upgrading XO to commit aa490
-
@JamfoFL As the bug appears to be non-trivial, I’ll create the issue on our end, and then we’ll see with the team to schedule this task. We’ll keep you updated here.
-
@MathieuRA
Yes, All running hosts and all running VMs -
B Bastien Nollet referenced this topic on
-
@MathieuRA Thanks so much! I appreciate all the effort!
-
@ph7 said in Lots of performance alerts after upgrading XO to commit aa490:
@olivierlambert
One strange note when I get the alarms on e-mail during a full backup of my running XOAlert: vm cpu usage > 65%, XO 67,8%
End of alert: vm cpu usage > 65% SyncMate 0,3%
End of alert is the wrong VM, should be XO.
In Dashboard/Health it's OK, it says XO in both alert and end of alert ??Not a big deal for me, but I thought might as well report it
I added a new VM and put it in a backup job (which had a concurrency of 2) and got these new alerts:
ALERT: VM CPU usage > 80% XO5: 80.4%
End of alert: vm cpu usage > 80% HomeAssistant: 6,7%
It seems like the reporting has changed and is reporting the latest VM.The host alert is reporting in the same way
ALERT: host memory usage > 90% X2 🚀: 92.7% used
END OF ALERT: host memory usage > 90% X1: 11.3%
I have changed all jobs to concurrency 1 and I have not got any SR alert
-
-
@Bastien-Nollet
Hi
I updated my test server and a copy of a Ronivay XO1st run
Host:
I only got 1+1 alert on e-mail, that seems OK.
2nd run
This time NO alert.
Maybe You should check some more
I will test again tomorrow, Good night... -
@ph7
And it did report the correct VM at the end of alert -
@ph7
3rd runNo alert.
Somehow, the CPU usage is lower.
I shall lower the alert limits to see what happens.4th run with 55% limit
I got 1+1 alarms this time.5th run
I got 1+1 alarms this time to.6th run
Increased the concurrency to 2.
this did put some more stress on the system.
only 1+1 alerts again.
Seems like You nailed it.Maybe You should check some more
And I am sorry for that, I was wrong
-
@Bastien-Nollet
More problems
I previous had 2 XO disabled from the alerts.But when I decided to remove both of them from the exclude or if I toggle the green
exclude VMs
"switch", I got this error:plugin.configure { "id": "perf-alert", "configuration": { "baseUrl": "Removed", "hostMonitors": [ { "smartMode": true, "alarmTriggerLevel": 60, "alarmTriggerPeriod": 60 }, { "smartMode": true, "variableName": "memoryUsage", "alarmTriggerLevel": 90 } ], "toEmails": [ "Removed" ], "vmMonitors": [ { "smartMode": true, "alarmTriggerLevel": 60, "alarmTriggerPeriod": 60, "excludeUuids": false, "uuids": [ "2f819438-5ab1-7309-99b0-7116313a03fe", "ee56c3c7-fd7d-c9aa-b2e1-ee698267e241" ] } ], "srMonitors": [ { "smartMode": false, "uuids": [], "excludeUuids": false } ] } } { "code": 10, "data": { "errors": [ { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf/0/not", "keyword": "not", "params": {}, "message": "must NOT be valid" }, { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf/1/const", "keyword": "const", "params": { "allowedValue": false }, "message": "must be equal to constant" }, { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf", "keyword": "anyOf", "params": {}, "message": "must match a schema in anyOf" }, { "instancePath": "/vmMonitors/0/uuids", "schemaPath": "#/properties/vmMonitors/items/oneOf/1/properties/uuids/not", "keyword": "not", "params": {}, "message": "must NOT be valid" }, { "instancePath": "/vmMonitors/0/excludeUuids", "schemaPath": "#/properties/vmMonitors/items/oneOf/2/properties/excludeUuids/const", "keyword": "const", "params": { "allowedValue": true }, "message": "must be equal to constant" }, { "instancePath": "/vmMonitors/0", "schemaPath": "#/properties/vmMonitors/items/oneOf", "keyword": "oneOf", "params": { "passingSchemas": null }, "message": "must match exactly one schema in oneOf" } ] }, "message": "invalid parameters", "name": "XoError", "stack": "XoError: invalid parameters at invalidParameters (/opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-common/api-errors.js:26:11) at default._configurePlugin (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/plugins.mjs:175:13) at default.configurePlugin (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/plugins.mjs:199:16) at Xo.configure (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/api/plugin.mjs:12:3) at Task.runInside (/opt/xo/xo-builds/xen-orchestra-202503081124/@vates/task/index.js:175:22) at Task.run (/opt/xo/xo-builds/xen-orchestra-202503081124/@vates/task/index.js:159:20) at Api.#callApiMethod (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/api.mjs:469:18)" }
I then tried to remove 1 VM , OK
Removed the 2nd one, OK
But when I tried to toggle the green switch I got this errorplugin.configure { "id": "perf-alert", "configuration": { "baseUrl": "Removed", "hostMonitors": [ { "smartMode": true, "alarmTriggerLevel": 60, "alarmTriggerPeriod": 60 }, { "smartMode": true, "variableName": "memoryUsage", "alarmTriggerLevel": 90 } ], "toEmails": [ "Removed" ], "vmMonitors": [ { "smartMode": true, "alarmTriggerLevel": 60, "alarmTriggerPeriod": 60, "excludeUuids": false, "uuids": [] } ], "srMonitors": [ { "smartMode": false, "uuids": [], "excludeUuids": false } ] } } { "code": 10, "data": { "errors": [ { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf/0/not", "keyword": "not", "params": {}, "message": "must NOT be valid" }, { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf/1/const", "keyword": "const", "params": { "allowedValue": false }, "message": "must be equal to constant" }, { "instancePath": "/vmMonitors/0/smartMode", "schemaPath": "#/properties/vmMonitors/items/oneOf/0/properties/smartMode/anyOf", "keyword": "anyOf", "params": {}, "message": "must match a schema in anyOf" }, { "instancePath": "/vmMonitors/0/uuids", "schemaPath": "#/properties/vmMonitors/items/oneOf/1/properties/uuids/not", "keyword": "not", "params": {}, "message": "must NOT be valid" }, { "instancePath": "/vmMonitors/0/excludeUuids", "schemaPath": "#/properties/vmMonitors/items/oneOf/2/properties/excludeUuids/const", "keyword": "const", "params": { "allowedValue": true }, "message": "must be equal to constant" }, { "instancePath": "/vmMonitors/0", "schemaPath": "#/properties/vmMonitors/items/oneOf", "keyword": "oneOf", "params": { "passingSchemas": null }, "message": "must match exactly one schema in oneOf" } ] }, "message": "invalid parameters", "name": "XoError", "stack": "XoError: invalid parameters at invalidParameters (/opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-common/api-errors.js:26:11) at default._configurePlugin (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/plugins.mjs:175:13) at default.configurePlugin (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/plugins.mjs:199:16) at Xo.configure (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/api/plugin.mjs:12:3) at Task.runInside (/opt/xo/xo-builds/xen-orchestra-202503081124/@vates/task/index.js:175:22) at Task.run (/opt/xo/xo-builds/xen-orchestra-202503081124/@vates/task/index.js:159:20) at Api.#callApiMethod (file:///opt/xo/xo-builds/xen-orchestra-202503081124/packages/xo-server/src/xo-mixins/api.mjs:469:18)" }
-
Some of the errors you encountered are intended. We don't allow values in the "Virtual Machines" field if "Exclude VMs" is disabled and "All running VMs" is enabled, because it would make the plugin configuration confusing.
However you're right, there seems to be an issue when the VMs are selected and then removed. The value becomes an empty list instead of being undefined, which causes the validation to fail when we try to turn off the "Exclude VMs" option.
I'm going to create a task on our side so that we can plan to resolve this problem.
In the meantime you can work around the problem by deleting the monitor and recreating a new one with the same parameters.