Repository navigation
|
Is anyone else having post update issues? I did an update in my lxc from 1.77.2 to 2.0.4 and it "works" but is extremely slow and laggy. The log in the container is being hammered with things like "info: Received tRPC request module="trpc" source="nextjs-react (json)" userId="undefined" userName="undefined"" So, clearly the db migration has issues. I've tried it a couple of times too. At least I can revert to 1.77... I tried a clean install with a new container and when I try to import my exported db I get the same type of log spam. I may just rebuild my dashboard from scratch, but I might wait until this disaster of an update works through some other bugs I've seen not even related to the above. |
Replies: 2 comments 4 replies
|
Well, I don't know where to go with this. I'm stumped. Clean install of the LXC and manually rebuilt board... and I'm still having a lot of timeouts. On the same Proxmox host I have an LXC with Docker. I installed Homarr using its native Docker install, imported my exact "broken" profile from the Community Scripts LXC and every single problem just vanished. Works PERFECTLY. The difference is astounds: Docker takes about 3 seconds to load my board with 30 items... the LXC? 20-30 seconds with many timing out. I'm left to assume there is something about the OS enviroment of the Community Scripts LXC that 2.0 is now incompatible with or causing excessive host test lag and timeouts. |
|
I've personally not had any issues with Homarr v2 or the migration, apart from a few widget issues that I've reported upstream. I also haven't noticed any significant increase in latency compared to v1. It's difficult to diagnose this without further information and logs. Given that you're also experiencing issues with the database migration, I'd suggest reporting this directly to the Homarr team. Our LXC script essentially deploys the upstream application, so there might be an underlying issue that needs to be addressed there. A few questions that might help narrow things down:
Without further details, it's kind of difficult to pinpoint a cause. Maybe @ajnart could also take a look at this. |
|
Appreciate the reply! I've tried this cleanly, fresh install with the copy/paste from the community script page on two Proxmox hosts including rebuilding a board from scratch. Same issue with no database import at all. If I restore a backup lxc to 1.77.2 on either of those hosts everything is fine. If I do an upgrade on that host, suddenly I get the problems I've noted. I have a Discussion over at the Homarr git with a few more things. In any of those cases the RAM/Resources stay about what they've always been back even before 1.7.x. Hardware should be more than enough on both hosts I've tried: an i7-8700 with 64GB RAM that generally runs at less than 10% CPU and a Core Ultra 5 245T with 64GB RAM that runs less than 15% CPU. I tried giving more resources to the lxc, but that didn't make a difference - not that they were using a fraction of what was assigned by default anyway. I've done all kinds of ping/curl/wget http/dns tests from the Homarr lxc to the most problematic services with timeouts and all are reachable and have 1-20ms responses (many are in the same host, just different containers.) There are also a couple of other folks in the Homarr Reddit sub who are seeing exactly what I'm seeing, which at least makes me think I'm not entirely crazy. lol |
|
@majorsl Hmm, weird then. Could you link those reddit threads here? Or possibly even better create an issue at homarr and link the reddit threads over there too? Thanks! |
|
I got it. I had one of those "wake up in the middle of the night" realizations. My Docker lxc is exempt from Adguard Home, most others are not. I have a rate limit on Adguard and Homarr 2.x must be much more efficient at requests. Bumped that way up and I'm back in business. It's always DNS. |
|
It's always DNS :D Glad you got it sorted out :) |
I got it. I had one of those "wake up in the middle of the night" realizations. My Docker lxc is exempt from Adguard Home, most others are not. I have a rate limit on Adguard and Homarr 2.x must be much more efficient at requests. Bumped that way up and I'm back in business.
It's always DNS.