View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000673 | Pgpool-II | Bug | public | 2020-12-16 18:07 | 2020-12-21 09:13 |
| Reporter | hao.wan@honeywell.com | Assigned To | pengbo | ||
| Priority | high | Severity | crash | Reproducibility | always |
| Status | feedback | Resolution | open | ||
| Product Version | 4.1.5 | ||||
| Summary | 0000673: cannot change into new db by using pgpool | ||||
| Description | Hi, please help have a look this issue. I did the following steps: step1 use pgpool to create database test1; step2 use \c test1, try to change into db test1, but returns error message: postgres=# \c test1 ERROR: unable to read message kind DETAIL: kind does not match between main(53) slot[1] (45) Previous connection kept attachment: log info 2020-12-16 16:59:16: pid 8207: DETAIL: connecting 0 backend 2020-12-16 16:59:16: pid 8207: DEBUG: creating new connection to backend 2020-12-16 16:59:16: pid 8207: DETAIL: connecting 1 backend 2020-12-16 16:59:16: pid 8207: DEBUG: authentication backend 2020-12-16 16:59:16: pid 8207: DETAIL: auth kind:0 2020-12-16 16:59:16: pid 8207: LOG: pool_read_kind: error message from 1 th backend:database "test1" does not exist 2020-12-16 16:59:16: pid 8207: ERROR: unable to read message kind 2020-12-16 16:59:16: pid 8207: DETAIL: kind does not match between main(53) slot[1] (45) 2020-12-16 16:59:19: pid 30554: DEBUG: authenticate kind = 0 2020-12-16 16:59:19: pid 30554: DEBUG: authenticate backend: key data received 2020-12-16 16:59:19: pid 30554: DEBUG: authenticate backend: transaction state: I 2020-12-16 16:59:19: pid 30554: DEBUG: authenticate kind = 0 2020-12-16 16:59:19: pid 30554: DEBUG: authenticate backend: key data received Is there any idea of this, thanks. | ||||
| Tags | No tags attached. | ||||
|
|
I would like to check the log of "step1 use pgpool to create database test1". Could you turn on the paramater "log_per_node_statement" and execute step1 and step2? log_per_node_statement = on Please share the entire pgpool log and pgpool.conf. |
|
|
I have checked the slave db, all the operations on the master, including add database and tables , do not duplicated on slave. if I use load_balance_mode , do I need to let the slave auto backup from master firstly? or it will sync the data from master to slave only by postgresql itself? I mean that, in this mode, does the database duplication operate by the pg or pgpool? |
|
|
> I mean that, in this mode, does the database duplication operate by the pg or pgpool? You can do both. Pgpool has several running modes. If PostgreSQL will sync the data from master to slave by postgresql itself, you need to use master_slave_mode: master_slave_mode = on master_slave_sub_mode = 'stream' If you want to replicate the data by using pgpool, you need to use replication_mode: replication_mode = on Running mode: https://www.pgpool.net/docs/41/en/html/runtime-config-running-mode.html Here is a configuration example using master_slave_mode. https://www.pgpool.net/docs/41/en/html/example-cluster.html |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2020-12-16 18:07 | hao.wan@honeywell.com | New Issue | |
| 2020-12-16 21:43 | pengbo | Assigned To | => pengbo |
| 2020-12-16 21:43 | pengbo | Status | new => assigned |
| 2020-12-16 21:49 | pengbo | Note Added: 0003667 | |
| 2020-12-16 21:49 | pengbo | Status | assigned => feedback |
| 2020-12-16 21:50 | pengbo | Note Edited: 0003667 | |
| 2020-12-17 10:28 | hao.wan@honeywell.com | Note Added: 0003672 | |
| 2020-12-17 10:28 | hao.wan@honeywell.com | Status | feedback => assigned |
| 2020-12-21 09:13 | pengbo | Note Added: 0003676 | |
| 2020-12-21 09:13 | pengbo | Status | assigned => feedback |
