[dev.icinga.com #11133] Icinga2 does unidexed deletes for servicegroup members

[icinga-migration]

This issue has been migrated from Redmine: https://dev.icinga.com/issues/11133

Created by vytenis on 2016-02-10 20:20:52 +00:00

Assignee: (none)
Status: Closed (closed on 2016-02-22 19:23:08 +00:00)
Target Version: (none)
Last Update: 2016-02-22 19:23:08 +00:00 (in Redmine)

Icinga Version: 2.4.1
Backport?: Not yet backported
Include in Changelog: 1

---

Config dumps slowed down a lot with git master at e04b708:

           Id: 1243139
         User: **
         Host: (***):61702
           db: ***
      Command: Query
         Time: 0
        State: updating
         Info: DELETE FROM icinga_servicegroup_members WHERE instance_id = 1 AND service_object_id = 106867
    Rows_sent: 0
Rows_examined: 50962

This is a table scan:

select count(*) from icinga_servicegroup_members;
+----------+
| count(*) |
+----------+
|    56812 |
+----------+
1 row in set (0.01 sec)

---

Relations:

• relates #11103

[icinga-migration]

Updated by vytenis on 2016-02-10 20:36:03 +00:00

Verified that:

ALTER TABLE `icinga_servicegroup_members` ADD INDEX `instance_obj_id` (`instance_id`, `service_object_id`);

made Icinga2 work again, instead of having infinite-length IDO DB queue.

[icinga-migration]

Updated by mfriedrich on 2016-02-22 19:22:42 +00:00

• Relates set to 11103

[icinga-migration]

Updated by mfriedrich on 2016-02-22 19:23:08 +00:00

• Status changed from New to Closed

Faulty fix in #11103, reverted it and therefore closing this issue.