-
Notifications
You must be signed in to change notification settings - Fork 81
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
pgcopydb seems to get stuck after setting endpos #427
Comments
I see same issue. After setting the
But receive and transform seem to be blocked in file write,
|
Few a times the apply process blocks forever when the follower switches to live replay mode,
IIUC, the apply process waits for the transaction metadata file
I could see that transaction commit at line # 38046 in 0000000400000119000000E7.json, but not in the transformed sql file. Refer the attached files below, files-0000000400000119000000E7.zip Also attached a full log. |
Hi, It's unclear to me if you're hitting twice the same problem or two different problems.
The second problem seems related maybe to your previous area of work with resuming logical decoding. The first problem is strange because we use |
@dimitri Both of the issues mentioned above [1] and [2] occur after
It seems the |
I see that transform process blocked in read_from_stream and we never check endpos being reached from transform. I think the easiest approach would be sending SIGTERM from EDIT: I see that there is a mechanism in place for transform, but it never gets called because the input stream for transform didn't had any content and it just loops around in |
@dimitri IIUC, this is caused by a deadlock during replay mode. The |
Let me create a separate issue to track deadlock while doing |
Not yet entirely sure whether this is a bug or I'm doing something wrong, but while doing a
pgcopydb clone --follow
and usingpgcopydb stream sentinel set endpos --current
, I regularly see pgcopydb getting (permanently) stuck.Log suggests pgcopydb does exit the loop startLogicalStreaming as expected:
Some stacktraces from different pgcopydb processes:
It happens both with writes and no writes during the clone.
Version:
The text was updated successfully, but these errors were encountered: