Hello everybody
We had some trouble with our Zabbix lately. We run it on a KVM virtualisation which runs on a GlusterFS cluster.
We occured some crashes and now our database is messed up.
What we already did:
- Remove the old database
- Import old scheme from backup
- Import data from backup
- Delete and recreate corrupted table
It looks like its working now, but the history tables are still not working properly.
That's very painful since the datareports of our objects can't be stored correctly and some errors apear, even tho there is no error on a client.
When i list all tables i see all the tables:
But when i am running a check on the history_uint i get this error:
Thats kinda strange, because the Zabbix server is writing new data in the table. (Our backup didn't contain history data)
We already tried to delete the history_uint table and created a new one, but the error still occures.
Does anybody have a solution or an idea how to fix this?
Thanks a lot!
We had some trouble with our Zabbix lately. We run it on a KVM virtualisation which runs on a GlusterFS cluster.
We occured some crashes and now our database is messed up.
What we already did:
- Remove the old database
- Import old scheme from backup
- Import data from backup
- Delete and recreate corrupted table
It looks like its working now, but the history tables are still not working properly.
That's very painful since the datareports of our objects can't be stored correctly and some errors apear, even tho there is no error on a client.
When i list all tables i see all the tables:
Code:
| history | | history_log | | history_str | | history_text | | history_uint |
Code:
MariaDB [zabbix]> check table history_unit; +---------------------+-------+----------+-------------------------------------------+ | Table | Op | Msg_type | Msg_text | +---------------------+-------+----------+-------------------------------------------+ | zabbix.history_unit | check | Error | Table 'zabbix.history_unit' doesn't exist | | zabbix.history_unit | check | status | Operation failed | +---------------------+-------+----------+-------------------------------------------+ 2 rows in set (0.001 sec)
Thats kinda strange, because the Zabbix server is writing new data in the table. (Our backup didn't contain history data)
Code:
MariaDB [zabbix]> select * from history_uint LIMIT 20; +--------+------------+-------+-----------+ | itemid | clock | value | ns | +--------+------------+-------+-----------+ | 50081 | 1681970321 | 0 | 502806857 | | 50081 | 1681974581 | 0 | 807556469 | | 50081 | 1681975121 | 0 | 40932666 | | 50081 | 1681975601 | 0 | 39593459 | | 50081 | 1681983461 | 0 | 49031482 | | 50081 | 1681985441 | 0 | 61489273 | | 50081 | 1681999901 | 0 | 615364904 | | 50081 | 1682003921 | 0 | 292049754 | | 50081 | 1682022641 | 0 | 146939188 | | 50081 | 1682033201 | 0 | 30275079 | | 50081 | 1682045501 | 0 | 21339457 | | 50081 | 1682045681 | 0 | 601833878 | | 50081 | 1682047181 | 0 | 495823686 | | 50081 | 1682052161 | 0 | 564115976 | | 50081 | 1682052881 | 0 | 451397606 | | 50112 | 1681969766 | 3 | 176889242 | | 50112 | 1681970246 | 3 | 546031608 | | 50112 | 1681970786 | 3 | 444488404 | | 50112 | 1681970846 | 3 | 204789835 | | 50112 | 1681971146 | 3 | 485798009 | +--------+------------+-------+-----------+ 20 rows in set (0.001 sec)
We already tried to delete the history_uint table and created a new one, but the error still occures.
Does anybody have a solution or an idea how to fix this?
Thanks a lot!
Comment