Replication appears to be down for Overpass & Tile Server

[jeffreyameyer]

Bug description Overpass is not getting updated with edits & neither is vtiles.

See: edits from changeset 83350 & afterward

Objects like Relation: Rancho Rincon de San Francisquito (2745225)

are not showing up in the vtiles:

Objects aren't reflected in Overpass downloads, either. : (

What should be happening?

Edits to the OHM db should be reflected immediately in Overpass and shortly thereafter in the tiles.

[batpad]

Digging into this now. @Rub21 does have a PR ready to run a job to monitor for missing replication files, but I don't feel comfortable deploying that without finding out the root cause of why we maybe missing replication files so often.

@geohacker if this is recurring missing replication files, may need you advice here on the best course of action / what we should be debugging.

[Rub21]

@jeffreyameyer @batpad , i found the issue, it is related to replication file

Update finished with status code: 3
2023-08-07 14:14:06 INFO: Using replication server at http://s3.amazonaws.com/planet.openhistoricalmap.org/replication/minute
2023-08-07 14:14:06 DEBUG: Using given sequence ID 1075902
2023-08-07 14:14:06 DEBUG: Starting download at ID 1075903 (max 100 MB)
2023-08-07 14:14:06 DEBUG: Error during diff download. Bailing out.
Empty version, skipping file

it is missing the sequence 1075903. looks like we need to merge https://github.com/OpenHistoricalMap/ohm-deploy/pull/222, to solve that

[batpad]

@Rub21 let's go ahead and merge that, but not close this issue until we have a good theory on why we're missing replication files this often. cc @geohacker

[Rub21]

@jeffreyameyer @batpad , I've just merged the monitoring script that is fixing the missing replication files too, Looks like now it is showing the boundary

also overpass is showing the boundary

[jeffreyameyer]

Overpass replication is currently down and the tile replication was down the other day.

Is this monitor working properly, or does it need to be updated / fixed to catch whatever happened when they both stopped?

Was Overpass also broekn by the corrupt replication file mentioned in #843 ?

[Rub21]

@jeffreyameyer the issues was because of the corrupted file , i fixed it, it is working now overpass-api.