qan won't show monngodb summary

qan won’t show monngodb summary, errors in client log: [TABLE]
[TR]
[TD]agent[/TD]
[TD]warning[/TD]
[TD]Cmd fail: Reply[Cmd:GetMongoSummary Error:‘exit status 1’ Id:48c76b9a-6ee9-495c-7cf6-2dafb31bf580][/TD]
[/TR]
[/TABLE]
[TABLE]
[TR]
[TD]agent[/TD]
[TD]err[/TD]
[TD]exit status 1[/TD]
[/TR]
[/TABLE]

and query profiles shows:
No data for selected time-range.

how can i run a debug process and find where is the resson causes this problem.

thanks.

Hi snowave,

is the latest version of Percona Toolkit installed on Client side?

yes , version is 3.0.4 , installed into /usr/bin

hm, can you share the output of the following commands?


env -i /usr/bin/pt-mongodb-summary
echo $?

as follows:

Instances

PID Host Type ReplSet Engine
16033 ip1:27020 PRIMARY repl1 wiredTiger
21238 ip2:27020 SECONDARY repl1 wiredTiger
31305 ip3:27020 SECONDARY repl1 wiredTiger

This host

Mongo Executable

Path to executable | /work/mongodb/bin/mongod

Report On ip1:27020

User | woker
PID Owner | mongod
Hostname | ip1:27020
Version | 3.2.10-3.0
Built On | Linux x86_64
Started | 2017-05-16 08:55:30 +0800 CST
Datadir | /work/mongodb/data/db_27020
Processes | 1
Process Type | replset

Running Ops

Type Min Max Avg
Insert 0 0 0/5s
Query 22 34 130/5s
Update 0 0 0/5s
Delete 0 0 0/5s
GetMore 327 364 1735/5s
Command 535 599 2854/5s

Security

Users : 58
Roles : 1
Auth : enabled
SSL : disabled
Port : 27020
Bind IP:

Cluster wide

Databases:

Balancer (per day)

Success:

hm, can you share output of the second command?

env -i /usr/bin/pt-mongodb-summary
echo $?

i want verify return code of pt-mongodb-summary tool

hi

echo $?

0

time pt-mongo-summary as follows:

real 0m8.061s
user 0m0.204s
sys 0m0.037s

so , i guess pt-mongo-summary timeout .

what PMM Server and PMM Client versions do you use?
Do you use http authentification? (SERVER_PASSWORD variable during “docker run …”)

pmm client version :pmm-client-1.2.0
server: via “docker pull percona/pmm-server:latest” command to fetch

the server doesn’t use a auth.

hi snowave ,

thank you for bug report, we found root of the issue.
it will be fixed in the next major release 1.3.0

very expect
we hope that we can use mongo queries in production environment as soon as possible.

it is possible to use mongo queries right now, but it is experimental and has many limitations at this moment.
if you want to check how it is working and report any bugs or ideas how to improve it – you can setup it according to document
https://www.percona.com/blog/2017/05/29/percona-monitoring-management-1-1-4-now-available/

thanks , i’ll have a try

hi ,when use pmm 1.3.1, click System Summary , next Mongo Summary , it shows :“cmd ‘Summary’ for subsystem ‘mongo’ failed: exit status 1”

exec pt-mongodb-summary

echo $?
0

can you run pt-mongodb-summary in the following way?

env -i /usr/bin/pt-mongodb-summary

Do you see any errors?

anyway PMM should be improved, I created a ticket for developers https://jira.percona.com/browse/PMM-1554

all print was correct ,

but the time of execute was long , as follows:

real 0m8.329s
user 0m0.266s
sys 0m0.025

so i think 8 seconds is bigger than wait timeout.

thanks.