Details
Minor
Description
When doint a select, at a random time, the select returns Can't create/write to file '/tmp/#sql_16d_0.MAI' (Errcode: 2 "No such file or directory"), and if retried it returns a "normal" result.
The issue occurs on 1 outta 3 master servers.
The table structure is :
> SHOW CREATE TABLE push; +-------+-------------+ | Table | Create Table | +-------+-------------+ | push | CREATE TABLE `push` ( `Id` int(11) unsigned NOT NULL AUTO_INCREMENT, `Token` varchar(200) DEFAULT NULL, `Visibility` int(11) DEFAULT '1', `BundleId` int(11) DEFAULT NULL, `Language` varchar(2) DEFAULT NULL, `Title` blob, `Content` blob, `isLocal` tinyint(1) DEFAULT '0', `Count` int(11) DEFAULT '0', `InstallId` int(11) DEFAULT NULL, `ReceiveDate` timestamp NULL DEFAULT NULL, PRIMARY KEY (`Id`), KEY `Token` (`Token`), KEY `bundle` (`InstallId`), KEY `visibility` (`Visibility`), KEY `ReceiveDate` (`ReceiveDate`), KEY `BundleId` (`BundleId`) ) ENGINE=InnoDB AUTO_INCREMENT=561491 DEFAULT CHARSET=utf8 | +-------+-------------+
Cluster is stable :
show status like "wsrep_clu%";
+--------------------------+--------------------------------------+
| Variable_name | Value |
+--------------------------+--------------------------------------+
| wsrep_cluster_conf_id | 194 |
| wsrep_cluster_size | 3 |
| wsrep_cluster_state_uuid | d13c363c-e443-11e4-a1fc-6b0c6015d8c4 |
| wsrep_cluster_status | Primary |
+--------------------------+--------------------------------------+
Gliffy Diagrams
Attachments
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Arnaudb : As pointed in the comment above, this could not be a bug.
Could you check the link above and see it its not a system issue?
It doesnt seem like there was a lack of space on this server, there was no state change in the cluster so no network flapping. It wasn't overload either.
To answer your second question, those server are in the same datacenter, not on the same physical host.
For starters, please check this discussion if you haven't yet: http://stackoverflow.com/questions/11997012/mysql-cant-create-write-to-file-tmp-sql-3c6-0-myi-errcode-2-what-does .
Maybe you'll find something relevant to your case.
Secondly, are your cluster nodes running on different physical/virtual machines, or on the same one?