# Celery task scheduler gets killed

**URL:** <https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985>\
**Category:** Administrators\
**Created:** [March 12, 2025, 10:15pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985 "2025-03-12T22:15:10Z")\
**Posts on this page:** 13\
**Page:** 1

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 12, 2025, 10:15pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/1 "2025-03-12T22:15:10Z")

</div>

Hi,

I am trying to install indico on a hosting service without sudo privileges. The installation seems to have gone well since the indico pace is up and running. The problem is that I continue to get a red prompt saying: \*\*The Celery task scheduler does not seem to be running. This means that email sending and periodic tasks such as event reminders do not work. \*\*.

I’ve tried to search for a solution, anyhow after several hours I did not find anything related to my problem. I’ve found hot that when executing:

```auto
indico celery worker -l DEBUG

```

I get the following output:

```auto

 -------------- celery@ --- v5.3.1 (emerald-rush)
--- ***** -----
-- ******* ---- Linux-4.18.0-553.30.1.lve.el8.x86_64-x86_64-with-glibc2.28 2025-03-12 23:00:52
- *** --- * ---
- ** ---------- [config]
- ** ---------- .> app: indico:0x7fedec691590
- ** ---------- .> transport: redis://127.0.0.1:6379/0
- ** ---------- .> results: redis://127.0.0.1:6379/0
- *** --- * --- .> concurrency: 12 (prefork)
-- ******* ---- .> task events: ON
--- ***** -----
 -------------- [queues]
                .> celery exchange=celery(direct) key=celery

[tasks]
  . category_cleanup
  . category_suggestions
  . celery.accumulate
  . celery.backend_cleanup
  . celery.chain
  . celery.chord
  . celery.chord_unlock
  . celery.chunks
  . celery.group
  . celery.map
  . celery.starmap
  . delete_unclaimed_files
  . event_reminders
  . heartbeat
  . indico.modules.attachments.tasks.generate_materials_package
  . indico.modules.events.static.tasks.build_static_site
  . registration_fields_retention_period
  . registration_retention_period
  . roombooking_end_notifications
  . roombooking_occurrences
  . send_email
  . static_sites_cleanup
  . survey_start_notifications
  . temp_cleanup
  . update_gravatars
  . update_system_notices

Killed

```

I am starting to think that this should not get killed… Does anyone have an idea on how to solve this issue ?

thank you in advance

---

<div class="post-metadata">

**Author:** ![ThiefMaster](https://talk.getindico.io/user_avatar/talk.getindico.io/thiefmaster/32/509_2.png) [@ThiefMaster](https://talk.getindico.io/u/ThiefMaster)\
**Post date:** [March 12, 2025, 10:17pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/2 "2025-03-12T22:17:20Z")

</div>

Out of memory maybe? This type of shared hosting service sounds like the kind of server that severely limits your resource usage as well.

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 12, 2025, 10:24pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/3 "2025-03-12T22:24:18Z")

</div>

maybe yes ? Actually it should have enough free memory:

 ![image](https://talk.getindico.io/uploads/default/original/2X/e/e189d7f424fea6f3943ef9462009b0e697627f6e.png)

I tried to look with `top` but it consumed only 0.3% of the available memory before getting killed.

---

<div class="post-metadata">

**Author:** ![ThiefMaster](https://talk.getindico.io/user_avatar/talk.getindico.io/thiefmaster/32/509_2.png) [@ThiefMaster](https://talk.getindico.io/u/ThiefMaster)\
**Post date:** [March 12, 2025, 10:26pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/4 "2025-03-12T22:26:37Z")

</div>

1 GB RAM is almost certainly too low to run Indico uWSGI w/ multiple processes and Indico Celery, likely also w/ multiple processes…

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 13, 2025, 8:18am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/5 "2025-03-13T08:18:01Z")

</div>

oh ! I’ll try to upgrade. Thank you!

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 13, 2025, 8:30am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/6 "2025-03-13T08:30:25Z")

</div>

Hi, any idea about this topic ? I have to upgrade my hosting plan, but I don’t know by hoe much. 2Gb is enough to manage a small non profit community ?

---

<div class="post-metadata">

**Author:** ![ThiefMaster](https://talk.getindico.io/user_avatar/talk.getindico.io/thiefmaster/32/509_2.png) [@ThiefMaster](https://talk.getindico.io/u/ThiefMaster)\
**Post date:** [March 13, 2025, 8:36am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/7 "2025-03-13T08:36:12Z")

</div>

You could also try tweaking your uwsgi config by setting `processes = 1` and `enable-threads = true`.

The celery worker also has an option to limit concurrent processes: `indico uwsgi worker -c 1`

That said, I recommend using tools like `top` or `htop` to check the memory usage of your processes.

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 13, 2025, 9:33am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/8 "2025-03-13T09:33:24Z")

</div>

Thanks again,

with `indico celery worker -c 1` it actually lasts longer but it gets killed anyway. It gets killed as soon as I reload the webpage. From top I don’t see the memory getting occupied, what I can see is that the CPU load increases.

I have no idea how to tweak the uwsgi config. Is it the file `etc/indico.conf`?

---

<div class="post-metadata">

**Author:** ![ThiefMaster](https://talk.getindico.io/user_avatar/talk.getindico.io/thiefmaster/32/509_2.png) [@ThiefMaster](https://talk.getindico.io/u/ThiefMaster)\
**Post date:** [March 13, 2025, 9:35am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/9 "2025-03-13T09:35:22Z")

</div>

> [@ipostuma](#):
>
> I have no idea how to tweak the uwsgi config. Is it the file `etc/indico.conf`?

No, `uwsgi-indico.ini`

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 13, 2025, 10:54am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/10 "2025-03-13T10:54:47Z")

</div>

Ah !

yeah, i had to skip that part of the installation procedure. I don’t have access to `/etc`. I suppose it is taken care of when the python App is started through the Cpanel.

This is the interface with what the user can manage:

 ![image](https://talk.getindico.io/uploads/default/original/2X/0/0ca6d92132654051fd36a7f714a22c7a754d71f4.png)

I have installed `pip install supervisor`, anyhow I didn’t manage to make it work.

---

<div class="post-metadata">

**Author:** ![ThiefMaster](https://talk.getindico.io/user_avatar/talk.getindico.io/thiefmaster/32/509_2.png) [@ThiefMaster](https://talk.getindico.io/u/ThiefMaster)\
**Post date:** [March 13, 2025, 10:57am UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/11 "2025-03-13T10:57:16Z")

</div>

In that case I have no idea how you tweak it, or what it even uses to run your indico application…

FWIW, this is not really a supported way to deploy Indico (and neither of us use cpanel), so I’m afraid there is not much we can do for you…

---

<div class="post-metadata">

**Author:** ![ipostuma](https://talk.getindico.io/user_avatar/talk.getindico.io/ipostuma/32/1817_2.png) [@ipostuma](https://talk.getindico.io/u/ipostuma)\
**Post date:** [March 13, 2025, 2:25pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/12 "2025-03-13T14:25:00Z")

</div>

I fully agree and understand, it is a total mess with cpanel and no privileges. I really didn’t think it could actually even display the interface. I had to connect the indico app to an aws postgre db through the MySQL port to make it work.

Anyhow, I would like to have a working indico app in the next months. As a non profit organization ([isnct.net](http://isnct.net)) we also might have enough budget to buy and maintain a small server. What would be the minimum requirements to make indico work well ? Considering that at peak usage we could have about 200/300 active users (I expect 10 daily users).

---

<div class="post-metadata">

**Author:** ![DeMiro5001](https://talk.getindico.io/user_avatar/talk.getindico.io/demiro5001/32/1701_2.png) [@DeMiro5001](https://talk.getindico.io/u/DeMiro5001)\
**Post date:** [March 14, 2025, 11:47pm UTC](https://talk.getindico.io/t/celery-task-scheduler-gets-killed/3985/13 "2025-03-14T23:47:14Z")

</div>

I am running one of my servers at home (using yunohost) and I am the maintainer of the indico package for yunohost (so you will find it in the catalog). If you have any machine with at least 4G of ram and a good internet connectivity, you can try it. Using yunohost you will get oneclick install, upgrade, backup, restore and remove. Let me know if you need other guidance. Good luck
