N
Novu4mo ago
AN1RUDH

Events keep getting merged with same old trigger event (Digest issue)

I've self-hosted novu (v0.24). It was working well previously. However, we've encountered an issue where trigger events are getting merged with a same (digest) event from past. (Please take a look at the video recording attached). I tried updating the digest schedule (can be seen in the video), but its still getting merged with the same exact (digest) event. I tried looking into logs (worker logs) but couldn't find anything useful. :/ Also, novu seems very slow (again, can be seen in the video). We've had M30-M40 tier mongo instance, yet we kept getting notifications form mongo for these --
Query Targeting: Scanned Objects / Returned has gone above 1000
Host has index suggestions
System: CPU (User) % has gone above 95
Query Targeting: Scanned Objects / Returned has gone above 1000
Host has index suggestions
System: CPU (User) % has gone above 95
The DB name created by default was test (not sure if this is expected). Lmk if I should post this in #⚓│community-self-host -- posted here as this seems to be unrelated to self-hosting. Thank you for your time 🙏
4 Replies
Pawan Jain
Pawan Jain4mo ago
@AN1RUDH The old digest event in which new events are getting merged is not completed yet If you change the digest window, it will take effect only once that old digest window gets completed I would recommend to you use cancel event api and cancel that digest event
AN1RUDH
AN1RUDHOP3mo ago
Thanks Pawan for responding. I tried this, but it didn't work. I think its because of this - https://github.com/novuhq/novu/blob/next/apps/api/src/app/events/usecases/cancel-delayed/cancel-delayed.usecase.ts#L62 It seems like its just updating the main digest. I still see the same issue.
Pawan Jain
Pawan Jain3mo ago
Ohh yes Looks like you will have to cancel all triggered events till now from starting
AN1RUDH
AN1RUDHOP3mo ago
Got it. Thanks for responding. I'll try and do that. I was able to do this. Seems to be working now. Thanks for the help. Also --
Also, novu seems very slow (again, can be seen in the video). We've had M40 (16GB RAM, 4vCPU) tier mongo instance, yet we kept getting notifications form mongo for these -- Query Targeting: Scanned Objects / Returned has gone above 1000 System: CPU (User) % has gone above 95 The DB name created by default was test (not sure if this is expected).
I wasn't expecting for mongo instance to struggle w/M40 config. Are there any optimizations that could help here? (I was thinking indices, but I'm not familiar w/Novu codebase that much so can't make a good decision there).
Want results from more Discord servers?
Add your server