2.6 x86 will not allow patches to install

2.6 x86 will not allow patches to install

Post by Devin L. Gang » Fri, 17 Jul 1998 04:00:00



SUMMARY OF PROBLEM
------------------
4 nearly-identical Solaris 2.6 Intel boxes all show the same errors when
trying to apply the latest recommended patch cluster.  The install_cluster
script shows that each patch installation is failing with an error code
of 5, and manual patch installation shows a problem with the checkinstall
script.  All four machines were installed with the same CD set.

DETAILS OF PROBLEM
------------------
I have four Solaris 2.6 Intel machines (dual P200MMX on Tyan Tomcat IV
motherboards, 128MB RAM) that I'm experiencing a problem with.  I
installed 2.6 on all of them with the same CD set, by the way, so I'm
starting to wonder if I might not have a flawed CD.  If that's the case,
I'm not sure what to do about it.

Yesterday, I downloaded the latest Recommended patch cluster and
attempted to install it and got errors similiar to this:

---------------------------------------------------------------------
Determining if sufficient save space exists...
Sufficient save space exists, continuing...
Installing patches located in /usr/local/patches/2.6_x86_Recommended
Using patch_order file for patch installation sequence
Installing 106126-05...
  Installation of 106126-05 failed. Return code 5.
---------------------------------------------------------------------

All patches in the cluster fail with the same return code, which when I
look it up in the installpatch script means "pkgadd failed."

Fine; the README depricates using installpatch in favor of patchadd.
I'll do that (even though that's what this version of installpatch is
really doing anyway...)

---------------------------------------------------------------------
# patchadd 106126-05

Checking installed packages and patches...
Verifying sufficient filesystem capacity (dry run method)...
Installing patch packages...
pkgadd: ERROR: checkinstall script did not complete successfully
---------------------------------------------------------------------

Great.  checkinstall is a script that comes with the patch.

So, I tried downloading the cluster again today, to see if there was
some bug in the version that was on the site.  No go.

Since I normally su to an account that's a root alias (to allow me to
take advantage of ksh and a more customized environment), I tried su'ing
to straight root, and got all of the same errors.

Does anyone out there have any clues why this might be happening, and
more importantly, how to fix it?  Three of the machines are not mission
critical, but one of them has been pressed into service for the time
being as our news server, and the software depends upon the presence of
certain patches in order to run without dumping core (or even better,
resetting or*) every 3 hours.

Please email as well as post replies, since like I said, my news server
is being affected and I might not get your responses if you just post.

Thanks in advance.

--

Chief Systems Administrator, Premier1 Internet Services, Sultan, WA, USA
    "Give me five Novaks, and _they_ could rule the world."
         -- Alistair J. R. Young

 
 
 

2.6 x86 will not allow patches to install

Post by Casper H.S. Dik - Network Security Engine » Fri, 17 Jul 1998 04:00:00


[[ PLEASE DON'T SEND ME EMAIL COPIES OF POSTINGS ]]


Quote:>---------------------------------------------------------------------
># patchadd 106126-05

>Checking installed packages and patches...
>Verifying sufficient filesystem capacity (dry run method)...
>Installing patch packages...
>pkgadd: ERROR: checkinstall script did not complete successfully
>---------------------------------------------------------------------
>Great.  checkinstall is a script that comes with the patch.

The solaris FAQ says:

6.54) Patch installation often fails with "checkinstall" errors.

    When installing a patch, the Solaris 2.5+ patch installation
    procedure will execute the script "checkinstall" with uid nobody.

    If any of the patch files or if any part of the path leading up to
    the patch directory cannot be read by nobody, an error similar to
    the following will appear:

    ./installpatch .
    Generating list of files to be patched...
    Verifying sufficient filesystem capacity (exhaustive method) ...
    Installing patch packages...
    pkgadd: ERROR: checkinstall script did not complete successfully
    ....

    You can workaround this in two ways, one is to make sure that
    nobody can read all patch files and execute a "pwd" in the patch
    directory or add an account "install" to /etc/passwd:

        install:x:0:1:installpatch braindamage:/:/bin/true

    Installpatch uses "nobody" as a fallback if it cannot find the "install"
    user.

    --- end of excerpt from the FAQ

Questions marked with a * or + have been changed or added since
the FAQ was last posted

The most recently posted version of the FAQ is available from
<http://www.wins.uva.nl/pub/solaris/solaris2/>
--
Expressed in this posting are my opinions.  They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.

 
 
 

2.6 x86 will not allow patches to install

Post by Clarence Wilkers » Fri, 17 Jul 1998 04:00:00


I had some problems with this kind of thing. The FAQ mentions
checking to see if all the directories in the path to the
patch directory have the right permissions.
--
Clarence Wilkerson      \ HomePage:     http://www.math.purdue.edu/~wilker

Dept. of Mathematics      \ Messages:   (765) 494-1903, FAX 494-0548
Purdue University,         \ Office:    (765) 494-1955 (voice/phonemail)
W. Lafayette, IN 47907-1395 \            Rm. 450 Math. Sci. Bldng.
 
 
 

2.6 x86 will not allow patches to install

Post by Dylan Neil » Sat, 18 Jul 1998 04:00:00


Here is my "never failed" method if intalling patche clusters ...

PS: I have Tyan Tomcat IV boards in ALL my Solaris Intel servers. They are
nice eh? :)

1. DOWNLOAD the Patch Cluster in .tar.Z format.
2. Copy the PATCH cluster to /
3. Reboot in Single User Mode (By saying "b -s" at the boot prompt).
4. cd /tmp
5. uncompress /<patchcluster.tar.Z>
6. While still in the /tmp directory ... tar -xvf /<patchcluster.tar>
7. ./install_cluster

This always works.. and some patches .. like the KERNEL patches .. which come
in the patch clusters _really should_ be installed in Single User.. this
isn't required, and on my Dual 200 Classic machines causes about 15 minutes
downtime .. but I'll live with that based on the fact that no rogue processes
are going to play with the Kernel while the patches are going in. This isn't
likely.. but certainly possible.

Make sure you uncompress and untar after the re-boot , as the /tmp directory
can be wiped by the OS as a matter of course.

Best of luck.

Dylan Neild
Speedline Canada Inc.


> SUMMARY OF PROBLEM
> ------------------
> 4 nearly-identical Solaris 2.6 Intel boxes all show the same errors when
> trying to apply the latest recommended patch cluster.  The install_cluster
> script shows that each patch installation is failing with an error code
> of 5, and manual patch installation shows a problem with the checkinstall
> script.  All four machines were installed with the same CD set.

> DETAILS OF PROBLEM
> ------------------
> I have four Solaris 2.6 Intel machines (dual P200MMX on Tyan Tomcat IV
> motherboards, 128MB RAM) that I'm experiencing a problem with.  I
> installed 2.6 on all of them with the same CD set, by the way, so I'm
> starting to wonder if I might not have a flawed CD.  If that's the case,
> I'm not sure what to do about it.

> Yesterday, I downloaded the latest Recommended patch cluster and
> attempted to install it and got errors similiar to this:

> ---------------------------------------------------------------------
> Determining if sufficient save space exists...
> Sufficient save space exists, continuing...
> Installing patches located in /usr/local/patches/2.6_x86_Recommended
> Using patch_order file for patch installation sequence
> Installing 106126-05...
>   Installation of 106126-05 failed. Return code 5.
> ---------------------------------------------------------------------

> All patches in the cluster fail with the same return code, which when I
> look it up in the installpatch script means "pkgadd failed."

> Fine; the README depricates using installpatch in favor of patchadd.
> I'll do that (even though that's what this version of installpatch is
> really doing anyway...)

> ---------------------------------------------------------------------
> # patchadd 106126-05

> Checking installed packages and patches...
> Verifying sufficient filesystem capacity (dry run method)...
> Installing patch packages...
> pkgadd: ERROR: checkinstall script did not complete successfully
> ---------------------------------------------------------------------

> Great.  checkinstall is a script that comes with the patch.

> So, I tried downloading the cluster again today, to see if there was
> some bug in the version that was on the site.  No go.

> Since I normally su to an account that's a root alias (to allow me to
> take advantage of ksh and a more customized environment), I tried su'ing
> to straight root, and got all of the same errors.

> Does anyone out there have any clues why this might be happening, and
> more importantly, how to fix it?  Three of the machines are not mission
> critical, but one of them has been pressed into service for the time
> being as our news server, and the software depends upon the presence of
> certain patches in order to run without dumping core (or even better,
> resetting or*) every 3 hours.

> Please email as well as post replies, since like I said, my news server
> is being affected and I might not get your responses if you just post.

> Thanks in advance.

> --

> Chief Systems Administrator, Premier1 Internet Services, Sultan, WA, USA
>     "Give me five Novaks, and _they_ could rule the world."
>          -- Alistair J. R. Young

 
 
 

1. tooltalk (rpc.ttdbserverd) problem after installing 2.6 x86 patch cluster

Administrators,

  Upon installing the 2.6 recommanded patch cluster for X86/Intel
I have begun to received errors in the messages system log from
rpc.ttdbserverd upon anyone logging in. The error message is as
follows:

Oct 23 08:31:03 npe-sol1 inetd[164]: /usr/dt/bin/rpc.ttdbserverd: Killed
Oct 23 08:37:02 npe-sol1 last message repeated 398 times

  The patch cluster includes Patch 105803-07 that changes
rpc.ttdbserverd, ttsession and libtt.so.2. But the log for the patch
indicated nothing was installed. Has anyone else experienced this
problem and if so how did you resolve it?  Should we back out the
patch? Is their a permission problem? Any suggestions?

Thanks in Advance,
Joel Whitesel

2. Sun stuff

3. Can not boot after install solaris x86(2.5 and 2.6) using AHA-1542CP.

4. Redhat 5.1 CDROM issue???

5. Solaris 2.6 x86 Patch Search

6. SLS+COMPAQ -- Unhappy To The Last?

7. cron x86 2.6 + y2k patches

8. PLEASE HELP! Got message: KERNEL PANIC: VFS: Unable to mount root fs on 02:1c

9. can't read Solaris 2.6 x86 hard disk on Solaris 2.6 Ultra-1 ?!?

10. HELP: 3C589D not recognized on Thinkpad 380ED (Solaris 2.6 x86)

11. x86 solaris 2.6 /dev/lp not working

12. Patched 2.6 3/98 and 2.6 5/98 the same?

13. Problem with x86 2.6 install