[PATCH] Fixes for NSS getnamebysid requests

[Jakub Hrozek]

Hi,

another problem I found while testing another patchset was that NSS
getnamebysid never consulted cache and went straight to BE. It was
because the cacheExpire attribute wasn't requested, so the default value
of 0 was used. Also negative cache wasn't consulted, but that's not so
much of a problem.

Attached are patches that fix both bugs and a unit test.

[Jakub Hrozek]

On Mon, Mar 30, 2015 at 06:59:07PM +0200, Jakub Hrozek wrote:
> Hi,
>
> another problem I found while testing another patchset was that NSS
> getnamebysid never consulted cache and went straight to BE. It was
> because the cacheExpire attribute wasn't requested, so the default value
> of 0 was used. Also negative cache wasn't consulted, but that's not so
> much of a problem.
>
> Attached are patches that fix both bugs and a unit test.

CI: http://sssd-ci.duckdns.org/logs/job/12/06/summary.html

[Lukas Slebodnik]

On (30/03/15 18:59), Jakub Hrozek wrote:
> Hi,
>
> another problem I found while testing another patchset was that NSS
> getnamebysid never consulted cache and went straight to BE. It was
> because the cacheExpire attribute wasn't requested, so the default value
> of 0 was used. Also negative cache wasn't consulted, but that's not so
> much of a problem.
>
> Attached are patches that fix both bugs and a unit test.

Tested with python module pysss_nss_idmap

ACK

LS

[Jakub Hrozek]

On Wed, Apr 22, 2015 at 07:36:20PM +0200, Lukas Slebodnik wrote:
> On (30/03/15 18:59), Jakub Hrozek wrote:
> > Hi,
> >
> > another problem I found while testing another patchset was that NSS
> > getnamebysid never consulted cache and went straight to BE. It was
> > because the cacheExpire attribute wasn't requested, so the default value
> > of 0 was used. Also negative cache wasn't consulted, but that's not so
> > much of a problem.
> >
> > Attached are patches that fix both bugs and a unit test.
>
> Tested with python module pysss_nss_idmap
>
> ACK
>
> LS

master:
    * bbd6f73bbad478a450ecfa2933a63de6dc269778
    * 6a074a5917a83c8414949b8c9c2b6d044bb652e6
    * ce6f3b6b2925d2c3ec02a76c3a1b6fbe4c7b145e

[Lukas Slebodnik]

On (24/04/15 16:49), Jakub Hrozek wrote:
> On Wed, Apr 22, 2015 at 07:36:20PM +0200, Lukas Slebodnik wrote:
> > On (30/03/15 18:59), Jakub Hrozek wrote:
> > > Hi,
> > >
> > > another problem I found while testing another patchset was that NSS
> > > getnamebysid never consulted cache and went straight to BE. It was
> > > because the cacheExpire attribute wasn't requested, so the default value
> > > of 0 was used. Also negative cache wasn't consulted, but that's not so
> > > much of a problem.
> > >
> > > Attached are patches that fix both bugs and a unit test.
> >
> > Tested with python module pysss_nss_idmap
> >
> > ACK
> >
> > LS
>
> master:
> bbd6f73bbad478a450ecfa2933a63de6dc269778
> 6a074a5917a83c8414949b8c9c2b6d044bb652e6
> ce6f3b6b2925d2c3ec02a76c3a1b6fbe4c7b145e

IMHO, at least 1st patch should be pushed to 1.12

LS

[Jakub Hrozek]

On Fri, Apr 24, 2015 at 05:35:58PM +0200, Lukas Slebodnik wrote:
> On (24/04/15 16:49), Jakub Hrozek wrote:
> > On Wed, Apr 22, 2015 at 07:36:20PM +0200, Lukas Slebodnik wrote:
> > > On (30/03/15 18:59), Jakub Hrozek wrote:
> > > > Hi,
> > > >
> > > > another problem I found while testing another patchset was that NSS
> > > > getnamebysid never consulted cache and went straight to BE. It was
> > > > because the cacheExpire attribute wasn't requested, so the default value
> > > > of 0 was used. Also negative cache wasn't consulted, but that's not so
> > > > much of a problem.
> > > >
> > > > Attached are patches that fix both bugs and a unit test.
> > >
> > > Tested with python module pysss_nss_idmap
> > >
> > > ACK
> > >
> > > LS
> >
> > master:
> > bbd6f73bbad478a450ecfa2933a63de6dc269778
> > 6a074a5917a83c8414949b8c9c2b6d044bb652e6
> > ce6f3b6b2925d2c3ec02a76c3a1b6fbe4c7b145e
>
> IMHO, at least 1st patch should be pushed to 1.12
>
> LS

OK, too bad the test patch doesn't apply:
    sssd-1-12: c61b3f050d82785525c845b839f545b5114b47f5