I doubt this may be related to 1197771 in general, this all looks quite different. But on the level of code I can't say.
"resolving two statements with the same ID" - I mean "updating rows with the same PK/UK id", but this is only a hypothesis, and this is definitely not the only requirement, there is something more.
With 5.5.37 we still have the issue, I was thinking to try the latest dev release (can't see when it's going to be released).
Collecting processlist seemed impossible last time we tried it if "WSREP: BF lock wait long" appeared, we will try once again.
And, btw, now we only have ONE "WSREP: BF lock wait long" in the log - when the bug was first met they go in series (as you can see in the attachment).
I doubt this may be related to 1197771 in general, this all looks quite different. But on the level of code I can't say.
"resolving two statements with the same ID" - I mean "updating rows with the same PK/UK id", but this is only a hypothesis, and this is definitely not the only requirement, there is something more.
With 5.5.37 we still have the issue, I was thinking to try the latest dev release (can't see when it's going to be released).
Collecting processlist seemed impossible last time we tried it if "WSREP: BF lock wait long" appeared, we will try once again.
And, btw, now we only have ONE "WSREP: BF lock wait long" in the log - when the bug was first met they go in series (as you can see in the attachment).