> Sounds like a systemd bug - systemctl should have some way of handling
> a potentially non-responsive server
We have no way of telling the difference between a "non-responsive
server" and a Redis instance that is writing its state to disk. This
is how Redis shuts down by design.
This also applies to sysvinit, it makes zero difference which init
system you are using from this point of view.
> Anyhow, the kill allowed the purge to complete
Well, sure, but now we don't know the real reason why it was hanging
for you. I would have liked to have seen the strace of the redis-server
process as that would have actually told us what it was doing, but
now you have killed it we will never know. :(
Regards,
--
,''`.
: :' : Chris Lamb
`. `'` la...@debian.org / chris-lamb.co.uk
`-
