View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000366 | Pgpool-II | Bug | public | 2017-11-13 15:55 | 2017-12-22 17:42 |
| Reporter | wdzj | Assigned To | |||
| Priority | urgent | Severity | major | Reproducibility | always |
| Status | closed | Resolution | open | ||
| Platform | Docker | OS | Centos | OS Version | 6 |
| Product Version | 3.6.6 | ||||
| Summary | 0000366: failed: FATAL: failed to read kind from backend | ||||
| Description | I also saw the two address content.I am using the 3.6.6 version, But there are still problems. There are 3 PostgreSQL 9.6 servers.Are the default configuration. I did not configure it? Tools and data are in the attachment. English is not good, hope can understand, thank you http://www.pgpool.net/mantisbt/view.php?id=271 https://git.postgresql.org/gitweb/?p=pgpool2.git;a=commitdiff;h=adcb63696b55463d6f6aa493b1b510e135ea67d7 | ||||
| Steps To Reproduce | Use osm2pgsql tool import osm data error shell: osm2pgsql -H localhost -U postgres -d opengis -s -S ./default.style ./bhutan-latest.osm | ||||
| Additional Information | osm2pgsql SVN version af61cae663 (64bit id space) release notes: 'Windows version built by Dominik Perpeet (http://www.customdebug .com/osm2pgsql/index.html)' Using projection SRS 900913 (Spherical Mercator) Unknown flag 'nocolumn' line 157, ignored Unknown flag 'nocolumn' line 158, ignored Unknown flag 'nocolumn' line 159, ignored Unknown flag 'nocolumn' line 160, ignored Unknown flag 'nocolumn' line 161, ignored Unknown flag 'nocolumn' line 162, ignored Setting up table: planet_osm_point NOTICE: table "planet_osm_point" does not exist, skipping NOTICE: table "planet_osm_point" does not exist, skipping NOTICE: table "planet_osm_point" does not exist, skipping NOTICE: table "planet_osm_point_tmp" does not exist, skipping NOTICE: table "planet_osm_point_tmp" does not exist, skipping NOTICE: table "planet_osm_point_tmp" does not exist, skipping WARNING: packet kind of backend 0 ['C'] does not match with master/majority nod es packet kind ['E'] ALTER TABLE planet_osm_point ALTER COLUMN way SET NOT NULL; failed: FATAL: failed to read kind from backend DETAIL: kind mismatch among backends. Possible last query was: "ALTER TABLE pla net_osm_point ALTER COLUMN way SET NOT NULL; " kind details are: 0[C] 1[E: column "way" of relation "planet_osm_point" does n ot exist] 2[E: column "way" of relation "planet_osm_point" message contents do not agree with length in message type "E" FATAL: failed to read kind from backend DETAIL: kind mismatch among backends. Possible last query was: "ALTER TABLE pla net_osm_point ALTER COLUMN way SET NOT NULL; " kind details are: 0[C] 1[E: column "way" of relation "planet_osm_point" does n ot exist] 2[E: column "way" of relation "planet_osm_point" message contents do not agree with length in message type "E" Error occurred, cleaning up | ||||
| Tags | No tags attached. | ||||
|
|
Tools and data |
|
|
Can you share pgpool.conf? |
|
|
pgpool.conf |
|
|
Why don't you use streaming replication mode? |
|
|
replication_mode = on This is replication mode? |
|
|
Yes, kind of. But these days users tend to prefer to use PostgreSQL streaming replication and set "master_slave = on" "master_slave_sub_mode = 'stream' (we call this "streaming replication mode") because it less restriction and more reliable than "replication_mode = on" (we call this "native replication mode"). |
|
|
I saw on the Internet,The master reading and writing,Slave read-only.I think everything can be read and write,So using replication mode. |
|
|
You have to be very careful not to mess up your database then. I am afraid your database has been already messed up. The error message indicates that table "planet_osm_point" does not exist on node 1 and 2. Or column "way" does not exist on node 1 and 2. Apparently your databases are not consistent. : "ALTER TABLE planet_osm_point ALTER COLUMN way SET NOT NULL; " kind details are: 0[C] 1[E: column "way" of relation "planet_osm_point" does n ot exist] 2[E: column "way" of relation "planet_osm_point" message contents do not agree with length in message type "E" |
|
|
Thank you for reminding me, use the test database now. :) Have a question to ask, "replication_mode = on” if a database outage, restart couldn't guarantee consistent data? |
|
|
Right. That's one of the reasons why I recommend streaming replication mode. |
|
|
May I close this issue? |
|
|
It can be turned off, and I've changed a pattern to avoid this problem. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2017-11-13 15:55 | wdzj | New Issue | |
| 2017-11-13 16:04 | wdzj | File Added: osm2pgsql.7z | |
| 2017-11-13 16:04 | wdzj | Note Added: 0001824 | |
| 2017-11-13 17:12 | t-ishii | Note Added: 0001825 | |
| 2017-11-13 17:16 | wdzj | File Added: pgpool.conf | |
| 2017-11-13 17:16 | wdzj | Note Added: 0001826 | |
| 2017-11-13 17:18 | t-ishii | Note Added: 0001827 | |
| 2017-11-13 17:20 | wdzj | Note Added: 0001828 | |
| 2017-11-13 17:25 | t-ishii | Note Added: 0001829 | |
| 2017-11-13 17:26 | wdzj | Note Added: 0001830 | |
| 2017-11-13 17:35 | t-ishii | Note Added: 0001831 | |
| 2017-11-13 17:40 | wdzj | Note Added: 0001832 | |
| 2017-12-08 11:17 | t-ishii | Note Added: 0001848 | |
| 2017-12-13 15:21 | t-ishii | Note Added: 0001864 | |
| 2017-12-13 15:22 | t-ishii | Status | new => feedback |
| 2017-12-18 10:04 | wdzj | Note Added: 0001869 | |
| 2017-12-18 10:04 | wdzj | Status | feedback => new |
| 2017-12-22 17:42 | t-ishii | Status | new => closed |
