BullMQ (with Redis) on Railway keeps getting ECONNREFUSED errors when running locally
I used the BullMQ Template that is offered by Railway, which spun up a node codebase for BullMQ and a redis server. I created a local version of the environment as well.
When I had a lot of jobs queued (maybe 100s), if I start the bullmq service locally it processes a few and then starts to get the following error message over and over - I also can't load the BullMQ UI.
[[ IP ADDRESS REMOVED BELOW]]
Any idea what's going on here? I would guess that Railway is stopping my connections from the local service, or redis is actually going down for some reason?
26 Replies
Project ID:
aa1a93fc-4414-42ac-9398-928082181803
You might find these helpful:
- I cannot connect to Redis
- read ECONNRESET on my database
- Can't connect to Postgres DB anymore
⚠️ experimental feature
aa1a93fc-4414-42ac-9398-928082181803
What powers this "You might find these helpful" service?
Hey! I wanted to follow up here to see if anyone had any recommendations on how I might debug this? I'm going to deploy to production and see if I hit similar issues, but wanted to see if this is a known Railway limitation before I explore changing implementation approaches
Hey team, just to follow up here so that the thread doesn't get left dangling.
What I did so far in trying to confirm the source of the
ECONNREFUSED
errors is to deploy to production and see if the running Railway service based on my Github repo has issue connecting to the production Redis instance running on Railway. At first, it didn't seem to have any issue connecting, but then as it continued to run, and I did another deploy to make my DB connection a shared singleton, I started seeing issues:all good, it appears that Discord running into a Slash command issue means that we couldn't raise the sev of the issue
I am tagging in @neb :) who can assist here
Thanks Angelo
The ERRCON reset is from the Redis?
Presumably
I am looking in:
aa1a93fc-4414-42ac-9398-928082181803
InterestingI don’t know what else it could be from. I do connect to a Postgres db on AWS but I don’t think that’s it
I was also connecting to a local db when trying it loacally
It works for a while then starts to econnrefused as I load jobs into bullmq
Oh
I am using the template service y’all had for it
I got it
Yea
I’m afk, will be back at my laptop in like 10 min, so might not be able to help on some issues
I’m also happy to hop on a call to debug together and share screen if helpful
Ah it looks like the next available time is a few days away
whoops
I need to deploy this service today, or else move off railway
throw any time on [email protected]
today
will clear the board for ya
I’ll be back in 15 minutes at mine - would that work?
that works!
Will ping you shortly, and thanks!
👋 hey @Angelo I'm free when you are - let me know what the best way to hop on would be
cool, spinning up a meet
hey none sec, not sure what's happening with Google Meet
It's stuck in Getting Ready...
Can I send you a Zoom?
sure thing!
Updating: seems to be a internal netoworking issue with our VMs and TCP. Working on a long term fix.
@Angelo I am also running into this issue, do you mind updating me as well when this is/will be fixed? 🙂
#🛂|readme - the Team isn’t online on the weekends and please don’t tag the team
I can look into it first thing my Monday
Great thanks!
FWIW my issue is fixed, I think it was something unrelated.