i just ran patchadd on 4 diff patches, and each time it said
that the packages were patched by patches w/ higher
rev numbers.
that got me to thinking -- if i could run patchadd as a
dry run, to determine which patches really needed
to be applied, i could save some downtime.
i already do a pre-screen of patches vs. installed patches.
but sometimes a patch (let's say) 111111-45 will be obsoleted
by 222333-78. my pre-screen won't catch that.
is there a way to catch those things?
thx.
j.
Unfortunately, Sun doesn't provide a tool that I know of to do what
you want. However, Patch Check Advanced (pca) should solve your
problem.
http://www.par.univie.ac.at/solaris/pca/
For example, 'pca -l missing' will display those patches which the
system requires. In your scenario, I believe that running 'pca
111111' will get you patch 222333 installed, since pca is that smart.
However, I haven't tried it, so I could be totally wrong. Check out
the pca usage page for more information.
http://www.par.univie.ac.at/solaris/pca/usage.html
Please post back with results!
doesn't help. i could get wget but pca says it covers recommended
and security. this came up in connection w/ live upgrade patches.
it appears that you can't get the source to patchadd. my
search of opensolaris returned 13 copies of a binary
license agreement.
List all the patches applied and grep for patch_id
# showrev -p | grep $PATCH_ID
Not sure what made you think that, but pca works with all patches, not
just R/S. What it can't do is to show dependencies for outdated patch
revisions, simply because this information isn't contained in Sun's
patchdiag.xref file.
Besides pca, maybe the undocumented "-a" option for patchadd is of help.
"patchadd -a 111111-45" will do a dry-run of the patch install, so
you'll see whether the patch can be successfully applied to a system or not.
Martin.
--
SysAdmin | Institute of Scientific Computing, University of Vienna
PCA | Analyze, download and install patches for Solaris
| http://www.par.univie.ac.at/solaris/pca/
the patchadd -a was what i needed. thx.
ahhhh... i looked through the top of the pca
program and what i saw there made it look like
it knew about those two kinds of patch clusters
and i jumped to the (i guess incorrect) conclusion
that those were what it was comparing against.
mea culpa maxima.
thanks again.