My original question:
> I have a real basic question about the patches that are
> distributed by Sun. If a patch has a higher version number,
> does it include all changes in patches with a lower version
> number? The reason I'm asking is because the README for patch
> 100075-06 (under "Problem Description") describes "Problems
> fixed by 100075-06 Patch" and "Problems fixed by 100075-05
> Patch", and the two lists are different. So, do I need to
> install 100075-05 and then 100075-06, or can I just install
> 100075-06 and get all the fixes?
Basically, most people felt that the highest version number of
a patch is the one that should be installed; not all the patches
up to and including the highest version of the patch. I have
included excerpts from some of the messages I received.
This all leads to another question. Maybe some of you can tell
me what you do when installing patches like this. If you
install a patch like 100075-05 and then 100075-06 comes along,
do you back off 100075-05 before installing 1000075-06 or do
you leave it installed? I ask this because I had previously
installed 100075-01 and it updated some different files than
the -05 and -06 patches. Should I have put back the old versions
of the files before installing -05 or -06?
I did not get any response on the net from Sun employees, so I
sent a message to our Sun tech. rep. She has since replied that
she is looking into it, but I haven't received a response from her.
In conclusion, I'm still kind of confused on the proper way to
install these patches.
RESPONSES:
From: era@niwot.scd.ucar.edu (Ed Arnold)
I'd appreciate a summary on this, cuz ... I had previously read
that a patch numbered nnnnnn-y ALWAYS supercedes a patch numbered
nnnnnn-x where y>x. If that's not true, it would be useful to
know.
From: ch@lks.csi.com
I don't know the answer to that but I do know that patches earlier to
100075-06 do not work very well (and some of them make everything much
worse -- like causing lots of kernel crashes). Patch 100075-06 has
fixed > 95% of our locking problems.
From: zjat02@trc.amoco.com (Jon A. Tankersley)
The first number of 100075-06 (100075) is some sort of patch ID number.
The second part is the version of that patch (06).
Here is a snap shot of what I have in my patch directory:
100040-01/ 100170-02/ 100224-01/ doit*
100075-05/ 100170-03/ 100224-02/ exabyte/
100075-06/ 100173-01/ 100228-02/ f77-1.3.1/
100092-03/ 100176-01/ 100232-01/ f771.2/
100098-01/ 100176-03/ 100280-01/ ingres/
100098-02/ 100179-01/ 12983.tar le0.patch.dir/
100100-01/ 100184-02/ 21363.tar.Z le0_patch/
100108-01/ 100185-01/ 24261.tar omni/
100121-03/ 100187-01/ 24895.tar portmap/
100125-02/ 100188-01/ 27035.tar.Z probs/
100125-03/ 100190-01/ 4.0.3/ psig.patch.tar
100133-01/ 100190-02/ 4.1/ shlib.etc.README
100138-02/ 100192-01/ 4.1.1/ tar_file
100141-01/ 100194-02/ OW2.0/ versa.sbus/
100141-02/ 100201-01/ README.sendmal virus/
You can see that there are two versions of 0098, 0141, 0170, 0190, and 0224.
I have split these into the 4.1, 4.1.1, 4.0.3 subdirs (that actually point right back up to these) to differentiate between the different OS's. Each patch fixes something different. Some are for F77, some are kernal patches,
some are OW patches, etc.
From: dal@gcm.com (Dan Lorenzini)
Normally you want to go with the latest patch because it usually
includes all the previous ones. In this case, I don't think you can
even get 100075-05 from Sun anymore now that 100075-06 is out.
But don't expect 100075-06 to solve all your lockd problems. We applied
it here and had our servers crash with "panic: iinactive". We have a
fix for that now and so far the server has not crashed, but file locking
is still less than perfect. I wish Sun would fix thing thing once and
for all. This is their sixth attempt and it works better but there are
still problems. Sheesh!
From: skipsun!skip@fsg.com (Skip Gilbrech)
I just looked at the README you mention and it is, indeed, ambiguous.
Last week I made up a complete list of patches on Sun's Support BB
and posted it (included below in case you missed it...).
[ deckel - I did not include this to save space - let me know if you
want a copy ]
>From the fact that 100075-05 doesn't appear in the list, I'd guess that
just the one patch (-06) is needed... If you find out from some
authoritative source, though, I'd like to hear the outcome..
-Skip
From: zjat02@trc.amoco.com (Jon A. Tankersley)
Neither. THey are different patches. You want the highest of the second
series of digits. 100075-05 is lower than 100075-06. They should contain
the same patches.
Debbie Eckel
Naval Surface Warfare Center
deckel@relay.nswc.navy.mil
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:14 CDT