[pca] synchronization between "recommended patch cluster" and patchdiag.xref?

[Ben Taylor]

On Tue, Oct 20, 2009 at 6:35 PM, Paul B. Henson <henson at acm.org> wrote:
>
> On a currently open support request, Sun is complaining that the patches on
> our server are way out of date and we should install the recommended patch
> cluster.

Did you send them an  explorer?

> We don't do patch clusters; we do pca :).
>
> A number of the patches they indicate should be installed do not appear to
> be marked as recommended in the patchdiag.xref (my current one is dated
> Oct/14/09), including for example:
>
>> ALERT: 140789 missing (current -01): SunOS 5.10_x86: nfsd, nfs4cbd, lockd patch
>
> 140789|01|May/18/09| | | |
> |10_x86|i386;118855-36;120012-14;127128-11;|SUNWnfscu:11.10.0,REV=2005.01.21.16.34;SUNWnfssu:11.10.0,REV=2005.01.21.16.34;|SunOS
> 5.10_x86: nfsd, nfs4cbd, lockd patch
>
>> ALERT: 140388 missing (current -01): SunOS 5.10_x86: statd patch
>
> 140388|01|May/14/09| | | |
> |10_x86|i386;118855-36;120012-14;|SUNWnfscu:11.10.0,REV=2005.01.21.16.34;|SunOS
> 5.10_x86: statd patch
>
>> ALERT: 141933 missing (current -02): SunOS 5.10_x86: unshare patch
>
> 141933|02|Aug/20/09| | | |
> |10_x86|i386;137138-09;141734-02;|SUNWnfssu:11.10.0,REV=2005.01.21.16.34;|SunOS
> 5.10_x86: unshare patch
>
> If they would have been marked as recommended, we would have already had
> them installed.

Since NFS isn't necessary to a machine, it's  no wonder it isn't recommended.

If you have NFS problems, then it's no wonder they're saying you're out of date.

I usually push back if some solution center yo-yo tries to tell me I
should be updating
patches that have nothing to do with the issue.  That's the solution
center's basic
mantra.  Update, update, update.  Which I can't fault them for.  It
those in-duh-viduals
which insist that some patch be installed which has no correlation to
the problem
at hand which raises my ire.

More times than I can count, some indeterminate problem has arisen on our
systems, and the solution center will tell me to patch, because they have no
more tools to debug the problem.  At that point, I'm usually talking
to backline.

> Has anyone compared the contents of the recommended patch cluster to the
> patchdiag file lately? I recall reading in the past occurrences where they
> were not the same.

With ZFS root, I'm just installing patches into a boot environment.  I
don't have
the cycles to match against recommended, and so I just let pca update to
current.  That also makes it easier with the solution center (many times I've
called in a problem and the first response is about patching.  So  nice to
say, updated as of "yesterday")

> It seems they really should be. Perhaps our resident Sun ambassador could
> provide some feedback :)?

Not really. Clusters are point in time "bundles", usually about every 4-6 weeks.
The patchdiag.xref file is constantly moving.  That's why I don't bother with
clusters or recommended.  patch to current, and I usually solve the issue of
solution center asking about patches.....  It's really a PITA when some solution
center engineer is holding up a real problem because of some perceive problem
with a patch not being there.