On Mon, Aug 28, 2017 at 06:25:17PM -0000, Corey Bryant wrote:
> Hi Felipe
>
> Historically we've configured OpenStack services to use sqlite at
> install time, and it makes sense to run the db sync commands in that
> case. However, I'm not sure why we also run the db sync commands if
> connection isn't defined (! grep -qE "^(sql_)?connection( )?=.*"
> /etc/nova/nova.conf).
>
> We do this in a number of packages so I want to figure out why we do
> that. We do the same in cinder, panko, aodh, nova, manila, to name a
> few.
>
> What we don't want to do is to run the db sync commands for non-sqlite
> db's. If not using sqlite we leave it up to the user or config mgmt
> tooling.
gotcha, then that's what that negation is meant for, the regexp needs to be tweaked a little bit to match only when "connection is config is set, but it's value contains sqlite"
On Mon, Aug 28, 2017 at 06:25:17PM -0000, Corey Bryant wrote: )?connection( )?=.*" nova.conf) .
> Hi Felipe
>
> Historically we've configured OpenStack services to use sqlite at
> install time, and it makes sense to run the db sync commands in that
> case. However, I'm not sure why we also run the db sync commands if
> connection isn't defined (! grep -qE "^(sql_
> /etc/nova/
>
> We do this in a number of packages so I want to figure out why we do
> that. We do the same in cinder, panko, aodh, nova, manila, to name a
> few.
>
> What we don't want to do is to run the db sync commands for non-sqlite
> db's. If not using sqlite we leave it up to the user or config mgmt
> tooling.
gotcha, then that's what that negation is meant for, the regexp needs to be tweaked a little bit to match only when "connection is config is set, but it's value contains sqlite"