db, node, sentry, sync: more exception handlers #2648
Conversation
canepat
added
the
maintenance
silkworm/db/snapshot_sync.cpp
Outdated
| SILK_WARN_M("SnapshotSync") << "SnapshotSync::run ex=" << ex.what(); | ||
| if (ex.code() == boost::system::errc::operation_canceled) { | ||
| SILK_WARN_M("SnapshotSync") << "SnapshotSync::run operation_canceled"; |
There was a problem hiding this comment.
| SILK_WARN_M("SnapshotSync") << "SnapshotSync::run ex=" << ex.what(); | |
| if (ex.code() == boost::system::errc::operation_canceled) { | |
| SILK_WARN_M("SnapshotSync") << "SnapshotSync::run operation_canceled"; | |
| SILK_ERROR_M("SnapshotSync") << "SnapshotSync::run ex=" << ex.what(); | |
| if (ex.code() == boost::system::errc::operation_canceled) { | |
| SILK_DEBUG_M("SnapshotSync") << "SnapshotSync::run operation_canceled"; |
operation_canceled could be a normal expected graceful exit, and error is an unrecoverable error in this case
P.S. if we implement #1527 this extra code could be removed
There was a problem hiding this comment.
Yep, but I need to have this warning to try and debug the root cause for issue #2333. I will remove these warnings later.
There was a problem hiding this comment.
@canepat can't you enable DEBUG level while debugging?
There was a problem hiding this comment.
@canepat can't you enable DEBUG level while debugging?
not in this case, because this is relevant during standalone sync and as you can see the exception handlers are in different modules (db, node, sentry, sync) so it would produce too many logs
There was a problem hiding this comment.
I've accepted your suggestion for one log and added TODOs for the other one
There was a problem hiding this comment.
I would either have this as a local patch or enable DEBUG and grep for "SnapshotSync::run operation_canceled" in the final log (where it exits). But if it makes your life easier, we can merge this temporarily until it is fixed.
No description provided.