email_events restarts repeatedly

[cmdntd]

When I install zulip by docker there are too many pid of email_events. It seems not to stop to complete
Could I wait for it? Or how to fix this?

zulip_1      | 2021-05-21 10:29:41,678 INFO success: zulip_events_email_senders entered RUN
NING state, process has stayed up for > than 1 seconds (startsecs)
zulip_1      | 2021-05-21 10:29:58,589 INFO exited: zulip_events_email_senders (exit status
 1; not expected)
zulip_1      | 2021-05-21 10:29:59,300 INFO spawned: 'zulip_events_email_senders' with pid
197
zulip_1      | 2021-05-21 10:30:00,309 INFO success: zulip_events_email_senders entered RUN
NING state, process has stayed up for > than 1 seconds (startsecs)

[timabbott]

What's happening here is that process is continuously crashing and then restarting. The crashes are likely caused by a configuration issue or potentially an OOM issue if your system have minimal memory; you should check its logs to see the actual traceback.

[WorldsEndless]

We are having the same issue on our side. Checking the internal logs for info....

[sp1z]

Just in case anyone else comes across this, it appears to be to do with not being able to communicate with the specified SMTP server correctly.

Make sure you have the following envs in docker-compose.yaml set correctly:

SECRETS_email_password
SETTING_ZULIP_ADMINISTRATOR
SETTING_EMAIL_HOST
SETTING_EMAIL_HOST_USER
SETTING_EMAIL_PORT
SETTING_EMAIL_USE_SSL
SETTING_EMAIL_USE_TLS

[bkraul]

As a sidenote, for me it was the SETTING_NOREPLY_EMAIL_ADDRESS setting. Basically on my email server configuration, the address must exist at least as an alias in order to be used as a sender address. Once I corrected that, it worked like a charm and the issue went away.

[mhassouna95]

We also use docker compose and encountered the same problem where the zulip_workers kept restarting (check with supervisorctl status). It turned out that rabbitmq was not working properly. After multiple fixing attempts we tried a newer version of rabbitmq (3.11.16 instead of 3.7.7) and that seems to have solved it.