influxdb: Debian package release v1.8.7 fails to start with systemd

Steps to reproduce: List the minimal actions needed to reproduce the behavior.

  1. wget https://dl.influxdata.com/influxdb/releases/influxdb_1.8.7_amd64.deb
  2. dpkg -i influxdb_1.8.7_amd64.deb
  3. systemctl restart influxdb.service

Expected behavior: Systemd is able to restart the service.

Actual behavior: Systemd fails to execute the script as user influxdb:

ls -la /usr/lib/influxdb/scripts/influxd-systemd-start.sh
-rwxr--r-- 1 root root  520 Jul 21 21:55 influxd-systemd-start.sh*

Environment info:

  • System info: Tested on Debian 9 stretch and 10 buster
  • InfluxDB version: v1.8.7

Logs:

Jul 23 11:57:56 phd-test-influx systemd[1]: Starting InfluxDB is an open-source, distributed, time series database...
Jul 23 11:57:56 phd-test-influx systemd[2853]: influxdb.service: Failed to execute command: Permission denied
Jul 23 11:57:56 phd-test-influx systemd[2853]: influxdb.service: Failed at step EXEC spawning /usr/lib/influxdb/scripts/influxd-systemd-start.sh: Permission denied
Jul 23 11:57:56 phd-test-influx systemd[1]: influxdb.service: Control process exited, code=exited, status=203/EXEC
Jul 23 11:57:56 phd-test-influx systemd[1]: influxdb.service: Failed with result 'exit-code'.
Jul 23 11:57:56 phd-test-influx systemd[1]: Failed to start InfluxDB is an open-source, distributed, time series database.
Jul 23 11:57:56 phd-test-influx systemd[1]: influxdb.service: Service RestartSec=100ms expired, scheduling restart.
Jul 23 11:57:56 phd-test-influx systemd[1]: influxdb.service: Scheduled restart job, restart counter is at 1.
Jul 23 11:57:56 phd-test-influx systemd[1]: Stopped InfluxDB is an open-source, distributed, time series database.

Steps to fix:

chmod +x /usr/lib/influxdb/scripts/influxd-systemd-start.sh
systemctl restart influxdb.service

About this issue

  • Original URL
  • State: closed
  • Created 3 years ago
  • Reactions: 19
  • Comments: 20 (5 by maintainers)

Most upvoted comments

Wouldn’t it be a good idea to release a fix in a new version as soon as possible? It’s really a very bad user experience that upgrading from 1.8.6 to 1.8.7 breaks a running installation (at least every Debian installation, and if I got the comment about RPM above right, also all RPM installations). Not every user will find the workaround in minutes. It’s unbelievable to me, that this issue is open for 9 days now without a fix being published nor 1.8.6-1 available in the repository for a downgrade…

Generally, it would be a good idea if older versions were still available in the repos. It would also help in staging scenarios where dev environments get an update while production environments still run a version that’s known to be running perfectly.

Same problem here. Would be great to see this fixed.

@Marci24h This was the solution for me. changed the max_attempts=10 to max_attempts=20 in /usr/lib/influxdb/scripts/influxd-systemd-start.sh

and then systemctl restart influxdb.service

Checked by systemctl status influxdb.service “running” instead of “starting”

thank you.

+1

you might have to fix the problems created by this after fixing with chmod +x

sudo chown -R influxdb:influxdb /var/lib/influxdb

1.8.10 includes a fix for this issue, and has been released.

same issue here after upgrade from 1.8.6-1 to 1.8.7-1 on debian 10 (buster) i need to make the start script executable. chmod +x /usr/lib/influxdb/scripts/influxd-systemd-start.sh i wrote about the issue on my small blog:

after upgrading to 1.8.9-1 the service stuck in a restart loop. but the executable issue was fixed 😉 also cause by the influxd-systemd-start.sh script. in line 25 a counter of 10 max attempts is spesified. since the loop sleeps for 1 second the timeout for starting the service is about 10 seconds.

since my installation is a bit bigger, it takes a little bit longer (14 seconds despite ssds). so i need to increase the value.

this problem is a bit similar to issue #21967 and #22110