Bug 809731

Summary: MySQL server: "innoDB: Error: unable to create temporary file" in /tmp when starting freshly installed mysql-server
Product: [Fedora] Fedora Reporter: Jirka Daněk <juraad>
Component: mysqlAssignee: Tom Lane <tgl>
Status: CLOSED RAWHIDE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: rawhideCC: chasesan, hhorak, tgl
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2012-04-07 08:18:21 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Jirka Daněk 2012-04-04 08:38:21 UTC
Description of problem:
When starting mysqld, the process fails:

systemctl start mysqld.service
Job failed. See system journal and 'systemctl status' for details.

tail /var/log/messages
Apr  4 10:27:59 private systemd[1]: mysqld.service holdoff time over, scheduling restart.
Apr  4 10:27:59 private systemd[1]: Job pending for unit, delaying automatic restart.
Apr  4 10:27:59 private systemd[1]: Unit mysqld.service entered failed state.
Apr  4 10:28:00 private mysqld_safe[5085]: 120404 10:28:00 mysqld_safe Logging to '/var/log/mysqld.log'.
Apr  4 10:28:00 private mysqld_safe[5085]: 120404 10:28:00 mysqld_safe Starting mysqld daemon with databases from /var/lib/mysql
Apr  4 10:28:00 private systemd[1]: mysqld.service: main process exited, code=exited, status=1
Apr  4 10:28:00 private systemd[1]: mysqld.service: control process exited, code=exited status=1
Apr  4 10:28:00 private systemd[1]: mysqld.service holdoff time over, scheduling restart.
Apr  4 10:28:00 private systemd[1]: Job pending for unit, delaying automatic restart.
Apr  4 10:28:00 private systemd[1]: Unit mysqld.service entered failed state.

tail -n50 /var/log/mysqld.log 
120404 10:27:59 mysqld_safe Starting mysqld daemon with databases from /var/lib/mysql
120404 10:27:59 [Note] Plugin 'FEDERATED' is disabled.
120404 10:27:59 InnoDB: The InnoDB memory heap is disabled
120404 10:27:59 InnoDB: Mutexes and rw_locks use GCC atomic builtins
120404 10:27:59 InnoDB: Compressed tables use zlib 1.2.5
120404 10:27:59 InnoDB: Using Linux native AIO
/usr/libexec/mysqld: Can't create/write to file '/tmp/ibko5VRn' (Errcode: 13)
120404 10:27:59  InnoDB: Error: unable to create temporary file; errno: 13
120404 10:27:59 [ERROR] Plugin 'InnoDB' init function returned error.
120404 10:27:59 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
120404 10:27:59 [ERROR] Unknown/unsupported storage engine: InnoDB
120404 10:27:59 [ERROR] Aborting

120404 10:27:59 [Note] /usr/libexec/mysqld: Shutdown complete


Version-Release number of selected component (if applicable):
mysql-server.x86_64 0:5.5.22-1.fc18

How reproducible, Steps to Reproduce:
Install mysql-server.x86_64 0:5.5.22-1.fc18 and try to start it using systemctl

Comment 1 Tom Lane 2012-04-04 13:32:27 UTC
Are you up to date on systemd and selinux-policy?

5.5.22 enables the PrivateTmp stuff, which is still new and probably buggy.  If it's not working I would suspect those components, particularly in view of the failure being an EACCES one.

Comment 2 Jirka Daněk 2012-04-07 08:18:21 UTC
Disabling PrivateTmp in the systemd config did not help.

Upgrading my system did help. I just fell behind a little bit. Thanks

Comment 3 Tom Lane 2012-04-07 14:40:59 UTC
OK, sounds like it was just one of the recently-fixed bugs in PrivateTmp support then.

Comment 4 Robert Maupin 2012-05-07 21:54:07 UTC
Thanks, this information was very helpful. Glad to see it got sorted out.