Fix redis pipeline created and not used

[iamlikeme]

Fixes a mistake introduced in #364 where a Redis pipeline was created but not used (ArqRedis is used directly).
The pipeline was there to prevent race conditions, so it should be used.

[codecov]

Codecov Report

Merging #374 (18e57ea) into main (430df07) will increase coverage by 0.12%.
The diff coverage is 100.00%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #374      +/-   ##
==========================================
+ Coverage   98.64%   98.76%   +0.12%     
==========================================
  Files           9       11       +2     
  Lines         957     1050      +93     
  Branches      179      198      +19     
==========================================
+ Hits          944     1037      +93     
  Misses          6        6              
  Partials        7        7              

Impacted Files	Coverage Δ
arq/jobs.py	98.15% <100.00%> (+0.01%)	⬆️
arq/__init__.py	100.00% <0.00%> (ø)
arq/cron.py	100.00% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 430df07...18e57ea. Read the comment docs.

[samuelcolvin]

LGTM, thanks so much.

[arturzx]

Hello, it is chance for new release with this fix? I updated to latest 0.25 and this update leads to random fake ResultNotFound errors (proabably race condition because I use 0.01 poll_delay).

[samuelcolvin]

@JonasKs do you think we can do a new release?

I can help with it on Tuesday hopefully. (Once we have the first alpha of pydantic v2 out)

[JonasKs]

Yes! 😊 I'll look over all PRs tomorrow and see if there's some issues I can resolve.

I'm available all week. 😊

[samuelcolvin]

Amazing, thank you.

[arturzx]

Yes! blush I'll look over all PRs tomorrow and see if there's some issues I can resolve.

I'm available all week. blush

Hello, any progress on this? :-) Thank you!

[blakehawkins]

Hi friends, we're also hitting spurious/transient arq.jobs.ResultNotFound errors in arq 0.25

Wondering if there's a recommended workaround and/or rough timeline for the next release fixing this?

All the best

[m-kuhn]

FWIW, I added the following in my requirements.txt to avoid the arq.jobs.ResultNotFound errors:

arq @ git+https:/samuelcolvin/arq@9109c2e59d2b13fa59d246da03d19d7844a6fa19

A release would be appreciated

[m-kuhn]

Hello, it is chance for new release with this fix? I updated to latest 0.25 and this update leads to random fake ResultNotFound errors (proabably race condition because I use 0.01 poll_delay).

It seems you are looking for a quick roundtrip. I find myself in a similar situation and I am wondering if there would be a more resource friendly solution than this "spin loop", using pubsub (possibly with keyspace notifications) on one end and using one of the blocking redis functions (block until a job is is waiting).

[KShah707]

FWIW, I added the following in my requirements.txt to avoid the arq.jobs.ResultNotFound errors:

arq @ git+https:/samuelcolvin/arq@9109c2e59d2b13fa59d246da03d19d7844a6fa19

A release would be appreciated

+1. A new release with this fix would be appreciated.

[aviau]

Any plan on a bugfix release @samuelcolvin ?

I think we are also getting the JobNotFound error after upgrading to 0.25.

[samuelcolvin]

v0.26.0b1 is released, please try it, I'll release v0.26 at the end of the week, see #441.