On 2.8.0-rc5 (and all previous versions of MongoDB), a healthy server looked like this:
> db.adminCommand({getLog: "startupWarnings" })
{ "totalLinesWritten" : 0, "log" : [ ], "ok" : 1 }
Now, a healthy 3.0.0-rc6 mongod can look like this:
"startupWarnings": {
"totalLinesWritten": 1,
"log": ["2015-01-24T17:58:54.554+0000 I CONTROL [initandlisten] "],
"ok": 1
},
... and a genuine warning looks like this:
"startupWarnings": {
"totalLinesWritten": 4,
"log": [
"2015-01-24T14:47:04.736+0000 I CONTROL [initandlisten] ",
"2015-01-24T14:47:04.736+0000 I CONTROL [initandlisten] ** WARNING: /sys/kernel/mm/transparent_hugepage/defrag is 'always'.",
"2015-01-24T14:47:04.736+0000 I CONTROL [initandlisten] ** We suggest setting it to 'never'",
"2015-01-24T14:47:04.736+0000 I CONTROL [initandlisten] "
],
"ok": 1
},
MMS escalates "there is a startup warning" to the main Hosts page in the UI if totalLinesWritten > 0, which means that this new behavior in 3.0.0-rc5 is falsely triggering this check.
We can adapt MMS if necessary, but this will leave users of older version of MMS OnPrem seeing spurious warnings in the UI.
- is related to
-
SERVER-16643 extend existing THP warning to THP defrag setting
-
- Closed
-