Pm2: Error: ESPIPE: invalid seek, read

[rup2701]

What's going wrong?

Following two errors repeatedly in the logs:

Error: ESPIPE: invalid seek, read
PM2 error: Error caught while calling pidusage

Stacktrace:
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 error: Error caught while calling pidusage
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 error: Error: ESPIPE: invalid seek, read
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 log: Stopping app:API id:0
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 log: App [API:0] exited with code [0] via signal [SIGINT]
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 log: pid=32 msg=process killed
2019-07-20 14:27:23 default[20190720t070820] 2019-07-20T14:27:23: PM2 log: PM2 successfully stopped

How could we reproduce this issue?

We followed steps from here: http://pm2.keymetrics.io/docs/usage/use-pm2-with-cloud-providers/
to run a NodeJS app on Google AppEngine.

Supporting information

$ pm2 report

Not able to run it on AppEngine/shell.

[italoiz]

I am also getting the same error in Google Cloud Run. @rup2701 did you manage to solve?

[JonasLaux]

@italoiz i had the same problem with cloud run. After hours of trying out stuff, i just deleted the Cloud Run instance and redeployed it. This did the trick.

[shaikhriyaz434]

same problem with the cloud run. Locally working fine but when deployed, the log is flooded with the error.

[jimsorock]

Same, haven't tried deleting the service but I have redeployed/replaced the service.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[neil-119]

I am getting the same issue on Google Cloud Run as well:

2020-06-17 15:14:01.536 CDT2020-06-17T20:14:01: PM2 error: Error caught while calling pidusage
2020-06-17 15:14:01.536 CDT2020-06-17T20:14:01: PM2 error: Error: ESPIPE: invalid seek, read

... repeating endlessly and filling up Stackdriver logs. Did anyone manage to fix this? I tried deleting and redeploying the service but it had no effect.

[max-mykhailenko]

Have the same problem in docker

[marcusflat]

I have the same problem, but looks that anyone found a solution =/

[lazharichir]

@neil-119 have you found any solution for that? The same problem with pm2 on Cloud Run using pm2-runtime server.js but not finding any solution anywhere!

[ishaarawy]

I am getting the same error also on Google Cloud Run

[luisguerra]

__I know this is not the best solution but it worked for me__, the thing is that pidusage package cannot read the memory and cpu percentage through the procfiles (which is the default method), I fixed by modifying the file on the pm2 project (See here) where its being called and added the param { usePs: true }

According to the documentation on https://www.npmjs.com/package/pidusage

By default, pidusage will use procfile parsing on most unix systems. If you want to use ps instead use the usePs option:

pidusage(pid, {usePs: true})

The solution is to run this command (in my case I added to my Dockerfile)), which is part of the process)

RUN sed -i 's/pidusage(pids, function retPidUsage(err, statistics) {/pidusage(pids, { usePs: true }, function retPidUsage(err, statistics) {/' {PATH WHERE YOU HAVE INSTALLED YOUR MODULES}/node_modules/pm2/lib/God/ActionMethods.js