can you show the screenshots of the job options w/ the exclude list ? and run the job in debug mode and get the files that generates.
yep - there's some other yum output as well that is captured. that should show if it's using an exclude list or not. also what version of blade ?
Not getting anything in /var/log/yum.log? would an analysis generate any logs? I tried turning the debug level up to 10 in the analysis job. We are running 8.5.01.301
On the job set the ‘debug_mode_enabled’ property to true
Run the job
On the appserver that ran the wit for the job under NSH/tmp/debug there will be a directory named like the job name and the timestamp of the run. that should have the yum output.
Oh – also – are you using a customized yum.conf (this is 8.5?)?
I am not using a custom yum.conf. Do you know what files you want to see (some of them are quite large).
Config time: 0.012
Yum Version: 3.2.22
Setting up Package Sacks
pkgsack time: 11.143
rpmdb time: 0.000
Setting up Update Process
Building updates object
up:Obs Init time: 0.276
up:simple updates time: 0.183
up:obs time: 0.010
up:condense time: 0.000
updates time: 1.522
--> Running transaction check
---> Package crash.x86_64 0:5.1.8-3.el5_11 set to be updated
---> Package cups.x86_64 1:1.3.7-32.el5_11 set to be updated
---> Package cups-libs.x86_64 1:1.3.7-32.el5_11 set to be updated
---> Package cups-libs.i386 1:1.3.7-32.el5_11 set to be updated
---> Package firefox.x86_64 0:31.6.0-2.el5_11 set to be updated
---> Package firefox.i386 0:31.6.0-2.el5_11 set to be updated
---> Package kernel.x86_64 0:2.6.18-404.el5 set to be installed
---> Package kernel-devel.x86_64 0:2.6.18-404.el5 set to be installed
---> Package kernel-headers.x86_64 0:2.6.18-404.el5 set to be updated
---> Package ksh.x86_64 0:20100621-24.el5_11 set to be updated
---> Package nss_db.x86_64 0:2.2-38.el5_11 set to be updated
---> Package nss_db.i386 0:2.2-38.el5_11 set to be updated
---> Package openssl.x86_64 0:0.9.8e-32.el5_11 set to be updated
---> Package openssl.i686 0:0.9.8e-32.el5_11 set to be updated
---> Package openssl-devel.i386 0:0.9.8e-32.el5_11 set to be updated
---> Package openssl-devel.x86_64 0:0.9.8e-32.el5_11 set to be updated
---> Package postgresql-libs.x86_64 0:8.1.23-10.el5_10 set to be updated
---> Package postgresql-libs.i386 0:8.1.23-10.el5_10 set to be updated
---> Package sendmail.x86_64 0:8.13.8-10.el5_11 set to be updated
---> Package setroubleshoot.noarch 0:2.0.5-7.el5_11 set to be updated
---> Package setroubleshoot-server.noarch 0:2.0.5-7.el5_11 set to be updated
---> Package tzdata.x86_64 0:2015b-1.el5 set to be updated
--> Finished Dependency Resolution
Dependency Process ending
Depsolve time: 1.237
Package Arch Version Repository
kernel x86_64 2.6.18-404.el5 repo 22 M
kernel-devel x86_64 2.6.18-404.el5 repo 5.9 M
crash x86_64 5.1.8-3.el5_11 repo 2.3 M
cups x86_64 1:1.3.7-32.el5_11 repo 3.1 M
cups-libs x86_64 1:1.3.7-32.el5_11 repo 197 k
cups-libs i386 1:1.3.7-32.el5_11 repo 201 k
firefox x86_64 31.6.0-2.el5_11 repo 67 M
firefox i386 31.6.0-2.el5_11 repo 66 M
kernel-headers x86_64 2.6.18-404.el5 repo 1.5 M
ksh x86_64 20100621-24.el5_11 repo 1.3 M
nss_db x86_64 2.2-38.el5_11 repo 747 k
nss_db i386 2.2-38.el5_11 repo 759 k
openssl x86_64 0.9.8e-32.el5_11 repo 1.5 M
openssl i686 0.9.8e-32.el5_11 repo 1.5 M
openssl-devel i386 0.9.8e-32.el5_11 repo 1.9 M
openssl-devel x86_64 0.9.8e-32.el5_11 repo 1.9 M
postgresql-libs x86_64 8.1.23-10.el5_10 repo 197 k
postgresql-libs i386 8.1.23-10.el5_10 repo 198 k
sendmail x86_64 8.13.8-10.el5_11 repo 638 k
setroubleshoot noarch 2.0.5-7.el5_11 repo 134 k
setroubleshoot-server noarch 2.0.5-7.el5_11 repo 1.2 M
tzdata x86_64 2015b-1.el5 repo 755 k
Install 2 Package(s)
Update 20 Package(s)
Remove 0 Package(s)
Total download size: 181 M
Is this ok [y/N]: Exiting on user Command
Debug Flag: ON
Analysis Options:List - update
Initialize: cleaning up previous result files
clean up complete
Running YUM from /var/tmp/stage/LinuxCatalog_2715_ltacgplum04...
YUM Command: includelist=`cat rpm-includes.lst| tr -d '
' | tr '
' ' '` && for i in 0 1 2 3 4 5 :; do echo n; done | blyum -c ./yum.conf -C update $includelist > "/var/tmp/stage/LinuxCatalog_2715_ltacgplum04/yum_analysis.res" 2> "/var/tmp/stage/LinuxCatalog_2715_ltacgplum04/yum.err.log"
YUM dry run done!
Generating rpm information...
RPM information generated.
Getting installed RPMs count....
Record Analysis logs in logfile: /var/tmp/stage/LinuxCatalog_2715_ltacgplum04/analysis_log.log
Based on the debug yum log you provided, it's not doing any excludes. The excludes should be listed in the yum.conf on an exclude= line. It's not there. Could be a bug, not sure yet... Can you also show a screenshot of your yum.conf tab in the Patching Job setup? I think that's what Bill was referring to when he asked if you were using a custom yum.conf.
If you browse the Kernel smartgroup you have created, do you see the kernel packages in there or not?
As for excluding kernels, be careful because you're missing the redhat-release as well, so if you update redhat-release without the kernel for example, /etc/redhat-release file will show a later version of Red Hat while the actual kernel in use will be older. You need to exclude more than just kernel*. See this Red Hat KB about it: https://access.redhat.com/solutions/10185
I also prefer not to use the NAME property to filter packages using a smartgroup, but rather the RPM_PACKAGE* one and the "is one of" operator. By specifying the exact names of packages you want to exclude (or include) you're sure not to include/exclude extras you didn't know existed or didn't think about.
I don't see anything wrong in your setup so far, but the log clearly showed it was not using any excludes, so something is bugged. Have you tried to recreate the patching job and re-select the patch smartgroup as an exclude? Make sure you close and re-open your console first too in case it's caching incorrect data somewhere (although I doubt it).
Based on what I see I would open a support ticket.
what about in the patch global config, on the 'redhat' tab ?
Sorry for the long silence but this is just getting weirder. I tried creating a new analysis job with just my kernel smart group excluded
Any RPM Where ??NAME?? starts with kernel
and it appeared to work. I then exluded a smart group with the following filter:
Any RPM Where ??RPM_PACKAGE*?? starts with postgres
Running the analysis with the 2 smart groups exluded, and it shows both the kernel and postgres rpms as missing. I then removed the postgres group, so that it just excluded the kernel group (the exact configuration that worked before), and now I'm seeing the kernel rpm as missing?
Instead of "Any RPM Where ??RPM_PACKAGE*?? starts with postgres", use the "is one of" operator with the complete name of the package (i.e. the name of the package is just the first part before the version, i.e. kernel, or kernel-headers, or kernel-devel. Instead of using a partial string match, use the full package name with a list, that way you're sure not to include or exclude something you didn't really intend to.
You should also run the tests in an ordered manner, against the same server always, so you can compare apples with apples. Also, just to make sure, between each test, save your job and close it, then re-open it to confirm you're using the right settings and that they saved. If you have the job results window opened for the job, close it and re-open it between tests to make sure this is not a console glitch (i.e. you modified the job, but the results window kept the old settings). I doubt this is the case, but you need to be thorough regardless.
The debug logs (yum.conf and yum_analysis.res) from each run would also help to troubleshoot the issue.
what version of blade is this - down to the build ? eg 8.5.01.301