Looks like that yesterday at around 14:23 our parchercache hosts started to experience an increase on used disk space.
https://grafana.wikimedia.org/render/dashboard-solo/file/server-board.json?refresh=1m&panelId=17&orgId=1&from=1539149866002&to=1539236266002&var-server=1005&var-network=eth0&width=1000&height=500&tz=UTC%2B02%3A00
After we settle back to normal operation:
[] The binlog purger should be stopped on pc1004**
[] The binlog purger should be stopped on pc1005**
[] The binlog purger should be stopped on pc1006**
[x] The binlog purger should be stopped on pc2004**
[x] The binlog purger should be stopped on pc2005**
[] The binlog purger should be stopped on pc2006**
[x] The tables in 'parsercache' schema should be truncated on pc2004
[] The tables in 'parsercache' schema should be truncated on pc2005
[] The tables in 'parsercache' schema should be truncated on pc2006
[x] max_binlog_size should be set back to 1048576000*
```
* there were a lot's of data purged already, but the innodb won't free up space automatically
** they're running inside screen named 'purge' (screen -r purge)
*** or chage that in config too
```