Ticket #701 (closed defect: fixed)
reactivate should not unnecsssarily spam .xsession-errors
Reported by: | jdreed | Owned by: | |
---|---|---|---|
Priority: | low | Milestone: | Natty Alpha |
Component: | -- | Keywords: | |
Cc: | Fixed in version: | ||
Upstream bug: |
Description
Currently, when we restart CUPS inside the chroot, it spams .xsession-errors with:
* Starting Common Unix Printing System: cupsd ...done. invoke-rc.d.debathena-orig: unknown initscript, /etc/init.d/cupsys not found.
This is annoying and possibly confuses users.
Change History
comment:3 Changed 13 years ago by jdreed
- Status changed from committed to development
debathena / trac-#701 / andersk 16:43 (Anders Kaseorg) debathena / svn / jdreed 22:46 (This zephyr does not necessarily reflec Can I have an explicit ack on r25019 please, to confirm that what I d to policy-rc.d is sane? debathena / svn / andersk 00:21 (Anders Kaseorg) Looks alright. So, sure, ACK.
comment:5 Changed 13 years ago by kaduk
We will need to revisit this for Natty, which wants us to use the service(8) utility. This may require more logic in the code.
+ schroot -r -c login-b27588d9-1c53-40b8-b129-3b3f183faa96 -u root -- invoke-rc.d cups start Rather than invoking init scripts through /etc/init.d, use the service(8) utility, e.g. service cups start Since the script you are attempting to invoke has been converted to an Upstart job, you may also use the start(8) utility, e.g. start cups
comment:6 Changed 13 years ago by jdreed
- Status changed from proposed to closed
- Resolution set to fixed
This moved into production now. It's unclear that we need to switch to service (i.e., it's unclear that it's a deprecation message vs a "You're doing it wrong" message). If we do want/need to switch to service, open another ticket.
Note: See
TracTickets for help on using
tickets.
It wasn't so much starting CUPS as the error about cupsys not being found, which confused us when we asked users to look for errors in their .xsession-errors, or when they briefly saw it in the tcsh-sucks xterm and assumed that something bad had happened.
This is because reactivate's policy-rc.d makes no effort to determine what the CUPS initscript is called.
Fixed in r25019, but someone should review this, in case there's weirdness about what policy-rc.d should and shouldn't do.