View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000399 | Pgpool-II | Bug | public | 2018-05-21 12:25 | 2018-05-21 14:57 |
| Reporter | masano | Assigned To | t-ishii | ||
| Priority | normal | Severity | major | Reproducibility | always |
| Status | resolved | Resolution | open | ||
| OS | CentOS | OS Version | 7.4.1708 (Core) | ||
| Product Version | 3.7.3 | ||||
| Target Version | 3.7.4 | ||||
| Summary | 0000399: Pgpool-II always regards '%P' as 0 even when primary node id = 1 down occurs failover | ||||
| Description | Two PostgreSQL are streaming replication, and Pgpool-II is (Master/Slave mode, sub mode = 'stream'. When Primary backend node id 0 stopped, Pgpool-II log and failover script said Pgpool-II starts failover_command and '%P' is 0. However, when Primary backend node id 1 stopped, according to the llog, '%P' is still 0. In this circumstansis, %P should be 1, right? | ||||
| Steps To Reproduce | 1. Prepare servers (node01, node02). 2. Install PostgreSQL into node01, node02 and setup Streaming Replication (node01 -> node02). 3. Install Pgpool-II into noded01 and setup Master / Slave mode. 4. Start pgpool. 5. Stop primary PostgreSQL (node01). 6. Failover occured. It's OK. 7. Now, node02 is Primary PostgreSQL. Create Standby PostgreSQL at node01. 8. Stop primary PostgreSQL (node01). 9. Failover occured, but, %P parameter is not 1 but 0. | ||||
| Additional Information | When primary node = 1, standby node = 0 and stop the standby node, %P is 1 as expectedly. OS: - CentOS 7.4.1708 (Core) * 2 Pgpool-II: - pgpool-II-pg96-3.7.3-1pgdg.rhel7.x86_64 - pgpool-II-release-3.7-1.noarch - pgpool-II-pg96-extensions-3.7.3-1pgdg.rhel7.x86_64 PostgreSQL: // yum install postgresql96* (from PGDG) - postgresql96-libs-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-server-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-devel-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-contrib-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-test-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-tcl-2.4.0-1.rhel7.x86_64 - postgresql96-debuginfo-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-tcl-debuginfo-2.3.1-1.rhel7.x86_64 - postgresql96-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-odbc-10.02.0000-1PGDG.rhel7.x86_64 - postgresql96-plperl-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-pltcl-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-plpython-9.6.9-1PGDG.rhel7.x86_64 - postgresql96-docs-9.6.9-1PGDG.rhel7.x86_64 | ||||
| Tags | No tags attached. | ||||
|
|
|
|
|
I think it's a known issue reported at: https://www.pgpool.net/pipermail/pgpool-general/2018-May/006150.html A patch was proposed here. https://www.pgpool.net/pipermail/pgpool-general/2018-May/006151.html Please try it out. |
|
|
Thanks. Patched pgpool-II 3.7.3 worked well. The value of %P was correctly assigned. I look forward to RPM release. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2018-05-21 12:25 | masano | New Issue | |
| 2018-05-21 12:25 | masano | File Added: pgpool.conf | |
| 2018-05-21 13:24 | t-ishii | Note Added: 0002027 | |
| 2018-05-21 13:25 | t-ishii | Assigned To | => t-ishii |
| 2018-05-21 13:25 | t-ishii | Status | new => feedback |
| 2018-05-21 14:27 | masano | Note Added: 0002028 | |
| 2018-05-21 14:27 | masano | Status | feedback => assigned |
| 2018-05-21 14:57 | t-ishii | Status | assigned => resolved |
| 2018-05-21 14:57 | t-ishii | Target Version | => 3.7.4 |
