[MDEV-26788] Does Maxscale 6.1.3 prevent the reassembly of user variables? Created: 2021-10-08 Updated: 2021-10-13 Resolved: 2021-10-13 |
|
| Status: | Closed |
| Project: | MariaDB Server |
| Component/s: | OTHER |
| Affects Version/s: | 10.5.12 |
| Fix Version/s: | N/A |
| Type: | Bug | Priority: | Major |
| Reporter: | Christophe DEVEL | Assignee: | Sergei Golubchik |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | Compatibility, regression | ||
| Environment: |
Debian 10 |
||
| Issue Links: |
|
||||||||
| Description |
|
Hello We see a slight difference between version 10.3.27-MariaDB-0 + deb10u1-log and version 10.5.12-MariaDB-1: 10.5.12 + maria ~ buster-log. – Reproduce with
– Test script
– case tested 1.2- host local(user 'cluster') with PROC DEFINER (admin@%) <=> mariadb (10.3.27) => OK >>>> MariaDB 10.5 by Mascale 6.1.3 2.2- host local(user 'cluster') with PROC DEFINER (admin@%) <=> maxscale 6.1.3 <=> cluster (10.5.12) => KO 2.3- host local(user 'admin') with PROC DEFINER (cluster@%) <=> maxscale 6.1.3 <=> cluster (10.5.12) => KO 2.4- host local(user 'admin') with PROC DEFINER (admin@%) <=> maxscale 6.1.3 <=> cluster (10.5.12) => KO >>>> MariaDB 10.5 directement node master 3.2- host local(user 'cluster') with PROC DEFINER (admin@%) <=> node master (10.5.12) => OK 3.3- host local(user 'admin') with PROC DEFINER (cluster@%) <=> node master (10.5.12) => OK 3.4- host local(user 'admin') with PROC DEFINER (admin@%) <=> node master (10.5.12) => OK Regardless of the user <=> definer couple:
Would Maxscale block user variables? 1) First pass
> '@myvar_from_nodeMaster', null, '@myvar_from_maxscale', null;
>'@myvar_from_nodeMaster', 654321, '@myvar_from_maxscale', null; >>>> MariaDB 10.5 by Mascale 6.1.3 avec le user admin
> @myvar_from_node7, null, @myvar_from_maxscale, null
> @myvar_from_node7, null, @myvar_from_maxscale, 123456 2) Second pass
>'@myvar_from_nodeMaster', 654321, '@myvar_from_maxscale', null;
> @myvar_from_node7, null, @myvar_from_maxscale, 123456 Although we always connect with the same user, the variables remain watertight between the 2 connection spaces ... Is there a way around this? Thank you in advance for your return |
| Comments |
| Comment by Christophe DEVEL [ 2021-10-08 ] |
|
Maybe a regression that had been fixed in Maxscale version 2.1.9 : |
| Comment by Sergei Golubchik [ 2021-10-13 ] |
|
Thanks. Looks like |