Percona-qan-api not starting in docker container

[Makeshift]

I followed the guide for setting up a docker container here: https://www.percona.com/doc/percona-server/LATEST/installation/docker.html

After setting up, I proceeded to set up clients and the metrics monitor is working fine. However, the query monitor is not. The output of

pmm-admin check-network

shows that the Query Analytics API is down.

* System Time
PMM Server | 2018-02-06 16:25:34 +0000 GMT
PMM Client | 2018-02-06 16:25:39 +0000 GMT
PMM Client to PMM Server Time Drift | OK

* Connection: Client --> Server
-------------------- -------
SERVER SERVICE STATUS
-------------------- -------
Consul API OK
Prometheus API OK
Query Analytics API DOWN

Connection duration | 249.326µs
Request duration | 565.454µs
Full round trip | 814.78µs


* Connection: Client <-- Server
-------------- ----- ----------------- ------- ---------- ---------
SERVICE TYPE NAME REMOTE ENDPOINT STATUS HTTPS/TLS PASSWORD
-------------- ----- ----------------- ------- ---------- ---------
mysql:metrics test 10.22.3.20:42003 OK YES -

After delving into the docker container itself, and trying to start

/usr/share/percona-qan-api

manually, I get the following output. I found the below command by running ps ax.

[root@42d4d3569b29 bin]# /usr/sbin/percona-qan-api -srcPath /usr/share/percona-qan-api/src -importPath github.com/percona/qan-api -runMode prod
ERROR: logging before flag.Parse: I0206 16:19:13.471093 8014 logutil.go:31] log: Config dev.conf not found in paths:
panic: Config dev.conf not found in paths:

goroutine 1 [running]:
log.Panic(0xc420173ea8, 0x1, 0x1)
/usr/lib/golang/src/log/log.go:326 +0xc0
github.com/percona/qan-api/config.init.0()
/builddir/build/BUILD/qan-api-1.7.0/src/github.com/percona/qan-api/config/config.go:50 +0xd2
github.com/percona/qan-api/config.init()
<autogenerated>:1 +0xdd
github.com/percona/qan-api/app/db/mysql.init()
<autogenerated>:1 +0x62
github.com/percona/qan-api/app/db.init()
<autogenerated>:1 +0x49
github.com/percona/qan-api/app/agent.init()
<autogenerated>:1 +0x7f
github.com/percona/qan-api/app.init()
<autogenerated>:1 +0x75
main.init()
<autogenerated>:1 +0x5a

I’m not sure what to do from this point, as I don’t seem to be able to start the qan api in the docker container. The container has more than enough space, and this is a very recent clean install.

[Makeshift]

Apologies, this is the guide I followed: https://www.percona.com/doc/percona-monitoring-and-management/deploy/server/docker.setting-up.html

[lorraine.pocklington]

Hello there, thank you for posting. Hopefully the PMM team will be able to take a look at your question very soon, but in the meanwhile just in case this helps you find the solution, I thought I would share a couple of blog posts for you to run through on troubleshooting and installation:
[LIST]
[]https://www.percona.com/blog/2018/01/17/troubleshooting-percona-monitoring-and-management-pmm-metrics/
[]https://www.percona.com/blog/2017/02/24/installing-percona-monitoring-and-management-pmm-for-the-first-time-2/
[/LIST]

[Makeshift]

Thanks Lorraine - I had a look at these and they helped confirm that the qan-api is down, but unfortunately there’s no information on why it wouldn’t be working in the container.

[roma.novikov]

Makeshift , What version of Pmm Client and Servers do you using?

[Makeshift]

roma.novikov The server version is the docker container percona/pmm-server:latest, and it was grabbed yesterday, so it’s 1.7.0

The client version is 1.7.0, according to pmm-admin --version.

[Makeshift]

Since I didn’t get a reply, I tried 1.6.1 instead and that works fine. Pretty sure the latest version of your docker container doesn’t work.

[roma.novikov]

Makeshift , this is strange. the docker was supposed to work in any cases, especially in a clean install.

Can you also check and run the tool from this tutorial https://www.percona.com/blog/2016/11/01/percona-monitoring-and-management-pmm-information-script/ and collect additional data. Thanks