[MXS-3304] binlogrouter can not manage GTID sequence number higher than 4294967295 Created: 2020-11-20 Updated: 2021-08-02 Resolved: 2021-08-02 |
|
| Status: | Closed |
| Project: | MariaDB MaxScale |
| Component/s: | binlogrouter |
| Affects Version/s: | 2.4.13 |
| Fix Version/s: | 2.5.0 |
| Type: | Bug | Priority: | Major | ||||||||||||||||||||||||||||||
| Reporter: | Yurii | Assignee: | markus makela | ||||||||||||||||||||||||||||||
| Resolution: | Won't Fix | Votes: | 0 | ||||||||||||||||||||||||||||||
| Labels: | None | ||||||||||||||||||||||||||||||||
| Environment: |
Ubuntu 18 MariaDB server:
Maxscale (reproduces on version 2.4.13 as well):
Binlogrouter config:
|
||||||||||||||||||||||||||||||||
| Description |
|
I think Maxscale binlogrouter can not manage GTID sequence number higher than 4294967295. Steps to reproduce:
After starting replica position is correct and not moving. 2. running statement on db-allc01 (mariadb server):
3. Gtid_IO_Pos possition on maxscale side has significantly changed to 18-5141-2387487339: 6682454634 - 4294967295 = 2387487339
After stopping and starting replication stream with such position - we will get broken replication:
+Expected behaviour: + MaxScale binlog router service can process GTID sequence number higher than 4294967295 without any limitations. |
| Comments |
| Comment by markus makela [ 2020-11-23 ] | |||
|
MaxScale 2.2 is EOL, can you reproduce this with MaxScale 2.5? | |||
| Comment by Yurii [ 2020-11-23 ] | |||
|
It was reproduces with 2.4.13. Same steps, same issue. | |||
| Comment by markus makela [ 2020-11-23 ] | |||
|
The binlogrouter was rewritten for 2.5 and it's possible the problem is already solved there. | |||
| Comment by Yurii [ 2020-11-23 ] | |||
|
I do not have any information about 2.5. As I know, 2.4.13 version is supported version. Can we have a fix in 2.4 and other supported version of MaxScale? | |||
| Comment by markus makela [ 2020-11-24 ] | |||
|
2.5 is the latest release of MaxScale and it is a supported version. Testing with 2.5 would help identify whether a workaround exists for this problem. | |||
| Comment by Yurii [ 2020-11-26 ] | |||
|
I can not reproduce the issue in 2.5.5
Should we expect fix for 2.4 ? | |||
| Comment by markus makela [ 2020-12-08 ] | |||
|
For now I'd recommend using 2.5 as a workaround to this problem until a fix for 2.4 is created. | |||
| Comment by markus makela [ 2021-08-02 ] | |||
|
Since 2.5 is out and doesn't suffer from this, I don't think this can be easily fixed in 2.4. I'll close this as Won't Fix and I'd recommend using 2.5. |