View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000080 | Pgpool-II | Bug | public | 2013-10-29 14:35 | 2013-11-09 08:06 |
| Reporter | jamessewell | Assigned To | t-ishii | ||
| Priority | high | Severity | major | Reproducibility | always |
| Status | resolved | Resolution | open | ||
| Platform | x86_64 | OS | Ubuntu | OS Version | 12.04.3 |
| Summary | 0000080: memqcache_auto_cache_invalidation is unsettable | ||||
| Description | Installed pgpool-II version 3.3.1 (tokakiboshi)from PGDG. When using the following config: memqcache_method = 'shmem' memqcache_auto_cache_invalidation = on memory_cache_enabled = on memqcache_auto_cache_invalidation is not set as shown: show pool_status; <snip> memqcache_auto_cache_invalidation | 0 If I test the functionality with a SELECT, then an UPDATE, then a SELECT then I see that cache invalidation is not happening. Cheers, James | ||||
| Steps To Reproduce | 1. Install 2. SHOW pool_status 3. Observe incorrect value for memqcache_auto_cache_invalidation | ||||
| Tags | No tags attached. | ||||
|
|
It seems SHOW pool_status did not show the correct value of memqcache_auto_cache_invalidation. It always showed the value as 0, however pgpool actually recognizes the value correctly and the functionality has worked. How do you find auto cache invalidation is not working? |
|
|
SELECT a from t; -- Result 1 UPDATE t set a = 2; SELECT a from t; -- Result 1 SELECT a,b from t; -- Result 2 I will retest to ensure I can still observe this result. |
|
|
Works for me. test=# create table t(a int,b int); CREATE TABLE test=# select a from t; a --- (0 rows) test=# insert into t values(1,2); INSERT 0 1 test=# select a from t; a --- 1 (1 row) test=# update t set a = 2; UPDATE 1 test=# select a from t; a --- 2 (1 row) test=# select a,b from t; a | b ---+--- 2 | 2 (1 row) |
|
|
I am happy to open a new case for this, it seems that the problem is to do with schemas. It looks like it's pretty sick though! I can cause an error given the following configuration: memcache_expire = 0 memcache_auto_cache_invalidation = on And running the following tests: SHOW search_path; search_path ---------------- "$user",public SELECT a from t; a --- 1 3 SELECT a from public.t; a --- 1 3 SELECT a from jill.t; a --- 1 3 UPDATE t SET a=2; UPDATE jill.t SET a=2; SELECT a from t; a --- 2 2 SELECT a from public.t; a --- 1 3 SELECT a from jill.t; a --- 1 3 SELECT * from t; a --- 2 2 SELECT * from public.t; a --- 2 2 SELECT * from jill.t; a --- 2 2 UPDATE t SET a=3; SET search_path = jill; SELECT a from t; a --- 2 2 SELECT a from public.t; a --- 1 3 SELECT a from jill.t; a --- 1 3 SELECT * from t; a --- 2 2 SELECT * from public.t; a --- 2 2 SELECT * from jill.t; a --- 2 2 |
|
|
Seems the case of FAQ 1.27: http://www.pgpool.net/mediawiki/index.php/FAQ#When_I_use_schema_qualified_table_names.2C_pgpool-II_does_not_invalidate_on_memory_query_cache_and_I_got_outdated_data._Why.3F |
|
|
My apologies! Can I request that the need for this is made a bit clearer for those installing from PGDG / other repo sources? I didn't notice it as I skipped the building from source section. My fault, but could be avoided! Cheers, James |
|
|
I think you can easily install it from the pgpool-II source code. download pgpool-II-3.3.1.tar.gz unpack it by tar xfz pgpool-II-3.3.1.tar.g cd pgpool-II-3.3.1/sql/pgpool-regclass make make install (might require root privilege) $ psql template1 =# CREATE EXTENSION pgpool_regclass; In addition to above, if you already create databases, you need to CREATE EXTENTION for those databases. See http://www.pgpool.net/docs/latest/pgpool-en.html#install for more details. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2013-10-29 14:35 | jamessewell | New Issue | |
| 2013-10-31 00:01 | t-ishii | Changeset attached | => pgpool2 master a9af0e72 |
| 2013-11-03 17:58 | t-ishii | Note Added: 0000351 | |
| 2013-11-03 17:58 | t-ishii | Assigned To | => t-ishii |
| 2013-11-03 17:58 | t-ishii | Status | new => assigned |
| 2013-11-03 17:59 | t-ishii | Status | assigned => feedback |
| 2013-11-04 08:05 | jamessewell | Note Added: 0000354 | |
| 2013-11-04 08:05 | jamessewell | Status | feedback => assigned |
| 2013-11-04 09:34 | t-ishii | Note Added: 0000355 | |
| 2013-11-04 09:35 | t-ishii | Status | assigned => feedback |
| 2013-11-06 12:20 | jamessewell | Note Added: 0000357 | |
| 2013-11-06 12:20 | jamessewell | Status | feedback => assigned |
| 2013-11-06 14:50 | t-ishii | Note Added: 0000358 | |
| 2013-11-06 14:51 | t-ishii | Status | assigned => feedback |
| 2013-11-06 14:53 | jamessewell | Note Added: 0000359 | |
| 2013-11-06 14:53 | jamessewell | Status | feedback => assigned |
| 2013-11-06 14:58 | jamessewell | Note Edited: 0000359 | |
| 2013-11-09 08:05 | t-ishii | Note Added: 0000365 | |
| 2013-11-09 08:06 | t-ishii | Status | assigned => resolved |