-
ofrnxmr
pigeons looks like the homeserver is down
-
DataHoarder
yeah, rip
-
plowsof
-
plowsof
DataHoarder a notice ^
-
DataHoarder
notices get posted for your eyes only :)
-
DataHoarder
"context deadline exceeded" is timeout
-
DataHoarder
it fails on my own element too
-
DataHoarder
stuff on
matrix.monero.social:8448/_matrix/key/v2/server answers but that's not the endpoint that element tries either
-
DataHoarder
weird
matrix.monero.social/_matrix/client/versions loads on browser tab, but same page doesn't load under element.io on same browser
-
DataHoarder
I think the DB is borked plowsof, sending any request with an Authorization header (for auth) makes it timeout
-
pigeons
I didn't have this open, I see the report of matrix down. I just tested talking in a room that is on it and got replies, and tested various other things and they all seem to work so I wasn't aware its not working. I'll take a closer look now and figure it out and add an alert for whatever it ends up being and let you know
-
plowsof
DataHoarder
-
DataHoarder
endpoints still timeout
-
DataHoarder
time=2026-08-24T23:40:21.172+02:00 level=ERROR msg="Could not join matrix room" server=matrix room=#p2pool-log:monero.social err="request error: Post \"
matrix.monero.social/_matrix/client…log:monero.social?via=monero.social\": context deadline exceeded"
-
DataHoarder
any post to the matrix endpoints with a Bearer token for the accounts timeout
-
DataHoarder
replicated via curl as well. stuff without the Authorization header return instantly fine
-
DataHoarder
same issues testing from other network perspectives on monero.social user, the homeserver answers at some point but operations still timeout with 30s+ per response
-
pigeons
Thanks, working on it
-
pigeons
I think a bad disk, checking backups before pulling it
-
DataHoarder
bad IO is consistent with sluggish DB and checking authenticated calls :)
-
DataHoarder
or almost full disk, would be my other non-exotic guess
-
pigeons
try it now. Opening a ticket for the disk replacement and adding more alerting so disk errors and authenticated user matrix errors would notify
-
DataHoarder
it instantly worked on its own
-
DataHoarder
seems a bit slow but returns in 1-2s instead of ms, which is doable
-
DataHoarder
probably load due to clients catching up
-
pigeons
ok i'll continue to check it out but at least we have something
-
DataHoarder
> panic: M_UNKNOWN (HTTP 502): Failed to fetch alias
-
DataHoarder
some backend errors now from your side, probably partial timeouts
-
DataHoarder
seems bridge finally joined after getting happy with timeouts
-
DataHoarder
transactions still slow, so messages still are being dropped
-
DataHoarder
-
br-m
<ofrnxmr:xmr.mx> What happened to dh messages
-
br-m
<ofrnxmr:xmr.mx> And my test
-
br-m
<jpk68:matrix.org> If you're talking about DMs, I don't have a convo with your monero.social account
-
br-m
<ofrnxmr:xmr.mx> Im talking about my convos, groupa etx
-
br-m
<ofrnxmr:xmr.mx> Nothing loads, everything shows "yesterday"
-
DataHoarder[m]
it's still catching up
-
DataHoarder[m]
some rooms are now receiving messages from yesterday via federation
-
DataHoarder[m]
also messages from today show before the messages from yesterday