Hello, thanks for sharing the hint,
I've tested 10 minutes, but it did not fix that.
Having more time waiting for timeout, I've run strace on wal sender once
more and now the patern is different (probably I've catch it first time
before it reaches issue point).
There are repeated failing attempt
On 3/11/18 6:02 μ.μ., Jerry Sievers wrote:
Aleš Zelený writes:
Hello,
we are suing logical replication on 10.4 and it now hangs. After
some timeout it is retarted again, replaying 18GB of data and then
hangs (while 7GB of wals remains to be proceeded).
Timeout...
Have a look at the 2 setti
Hello,
wal_sender_timeout did not help (from default 60s to 600s), but I haven't
tested receiver timeout since it was sender complaining and receiver loops
over ( recvfrom with "Resource temporarily unavailable" result, so i
thought it is waiting for master sender to provide some data), thanks fo
Aleš Zelený writes:
> Hello,
>
> we are suing logical replication on 10.4 and it now hangs. After
> some timeout it is retarted again, replaying 18GB of data and then
> hangs (while 7GB of wals remains to be proceeded).
Timeout...
Have a look at the 2 setting wal sender/receiver timeout and yo
Hello,
we are suing logical replication on 10.4 and it now hangs. After some
timeout it is retarted again, replaying 18GB of data and then hangs (while
7GB of wals remains to be proceeded).
The backlog of 18GB comes from a failed migration adding new table to
replication while replication user w
