0
Background:
- Storage problem (Error 28) occured this morning. It want away after remove some logs under tmporary dir. However i found queries still stuck in "old" status and keep running. Before fixed they running more then 5000 seconds, after fixed they running more then 10000 seconds.
What i have done:
- Checked 'interactive_timeout' is 8 hours.
- Trying to execute one slow query manually. It takes around 2 seconds to fetch results.
- Killed all old connections incase things became more complicated.
Questions:
- What happened to those old connections? Why they were stuck in the old status, still running so slowly?