I'm running an upgrade on a test site in preparation for my sites migration from 3.7.1 to 4.X. the step did note 'this may take a long time' however it has been running for well over 5 hours.
If I run query status this is what I get:
Time Elapsed: 00:396:46, State: Repair with keycache, Query: ALTER TABLE post ADD INDEX threadid_visible_dateline (threadid,visible,dateline,userid,postid)
The server is still showing load, which rises and falls, so I think it is churning. My POST table has 2.1m rows and is about 1.7GB.
Is this normal? I can leave it overnight, but I really don't want my production change to take this long. Is the command line option quicker? Or does that just avoid time outs?
Any mysql tweaks I can do to help?
If I run query status this is what I get:
Time Elapsed: 00:396:46, State: Repair with keycache, Query: ALTER TABLE post ADD INDEX threadid_visible_dateline (threadid,visible,dateline,userid,postid)
The server is still showing load, which rises and falls, so I think it is churning. My POST table has 2.1m rows and is about 1.7GB.
Is this normal? I can leave it overnight, but I really don't want my production change to take this long. Is the command line option quicker? Or does that just avoid time outs?
Any mysql tweaks I can do to help?
Comment