[CentOS] how best to rollback from a yum update?

[Jon Detert]

Suppose I want to upgrade a bunch of packages on a system, but in case the 
upgrade produces unexpected, undesired results, I want to be able to rollback 
the system to its original state.  What is the best way to do that?

Often, I won't have, or be able to find, packages for the current installed 
versions.  I.e. If I haven't upgraded postgres for 2 years, it may be that I 
can no longer find a package for the version I am currently running.

The 'rollback' feature of rpm and yum [1] looks promising, but I have a few 
concerns with it:

a) why isn't the --rolback switch described in either the man page for rpm or 
the output from --help?  This makes me think the --rollback switch is 
deprecated or otherwise unofficial.

b) how can I trim/purge the 'repackaged' rpms that by default go in 
/var/spool/repackage [2]?  I don't want to just leave them there forever, but 
if I trim/purge/delete them, how will the rpm database be updated to know that 
it can no longer rollback to the date that they were installed?

AtDhVaAnNkCsE

[1] http://www.freedomit.co.nz/kb-centos/enabling-yum/rpm-rollback
[2] http://www.linuxjournal.com/article/7034?page=0,1
-- 
Jon Detert
Sr. Systems Administrator
Infinity Healthcare
Milwaukee, Wisconsin
414-290-6759
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] when is the o.s. considered to be at a certain minor version? Or, is it safe to apply only certain package updates from the next release version?

[Jon Detert]

Two related questions about the minor release numbers (e.g. the 'x' in 5.x or 
6.x) :

1) What constitutes the o.s. being at a particluar minor release?  Typically, 
when you install you are getting a package set available from a specific minor 
release number.  But what minor release is the o.s. at if I just update the 
centos-release package, and no other package?  Typically, a 'yum update' is 
said to take your whole package set to the latest minor release.  But what 
minor release is the o.s. if you just update certain packages (instead of 
taking all avaailable updates)?

2) Can I apply package updates made in a minor release greater than my current 
release without detriment to the integrity of the o.s.?
E.g. Suppose all my packages are at versions delivered in v5.6.
Suppose also, that I have a package xyz installed, and that an update to it was 
made available in v5.8.
Can I upgrade xyz to the updated version from 5.8 without updating other 
packages (except for any dependencies xyz has) to the versions available in 
v5.8?

Thanks,
-- 
Jon Detert
Sr. Systems Administrator
Infinity Healthcare
Milwaukee, Wisconsin
414-290-6759
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Will CentOS 6.1 still be supported?

[Jon Detert]

No more updates for v6.1.  This is the RedHat/CentOS model.  Think of the minor 
version numbers as 'service packs', as opposed to releases.

For example, let X < Y.  Then, Everything fixed in 6.x is also fixed in 6.y.  
However, updates new to v6.y are not made available for v6.x.

- Original Message -
> From: "Edward Martinez" 
> To: "CentOS mailing list" 
> Sent: Wednesday, December 21, 2011 2:31:19 PM
> Subject: [CentOS] Will CentOS 6.1  still be supported?
> 
> Hello,
> 
> 
> Got a question, will  CentOS6.1 continue to be  supported with
> updates,
> or is it mandatory to upgrade to 6.2?
> ___
> CentOS mailing list
> CentOS@centos.org
> http://lists.centos.org/mailman/listinfo/centos
> 
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Monitoring services

[Jon Detert]

did anyone mention https://www.icinga.org/ ?  I'm a long-time nagios user, but 
just heard about it yesterday.  It is a fork of nagios, has a more modern web 
interface, and nagios plugins are compatible with it.  It looks/sounds good.  
Anyone have experience with it?

- Original Message -
> From: m...@tdiehl.org
> To: "CentOS mailing list" 
> Sent: Tuesday, November 29, 2011 9:53:54 AM
> Subject: Re: [CentOS] Monitoring services
> 
> On Tue, 29 Nov 2011, Rajagopal Swaminathan wrote:
> 
> > Greetings,
> >
> > On Tue, Nov 29, 2011 at 12:05 AM, Paul Heinlein
> >  wrote:
> >> On Sun, 27 Nov 2011, Kenneth Porter wrote:
> >>
> >>> What's available to remotely monitor services?
> >
> > I have deployed Zabbix successfully to remotely monitor about 240+
> > geographically distributed locations connected by ADSSL links (IOW,
> > no
> > fixed IP) for the second largest public transport corporations
> > (next
> > only to Germany) in India successfully.
> >
> > Perhaps, you may consider that.
> 
> Another possibility is http://sourceforge.net/projects/xymon/
> 
> Regards,
> 
> --
> Tom   m...@tdiehl.org Spamtrap address
> me...@tdiehl.org
> ___
> CentOS mailing list
> CentOS@centos.org
> http://lists.centos.org/mailman/listinfo/centos
> 
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] not using LVM for Linux VM guests?

[Jon Detert]

Hello,

- Original Message -
> From: "Russell Smithies" 
> To: "CentOS mailing list" 
> Sent: Wednesday, November 16, 2011 2:37:54 PM
> Subject: [CentOS] not using LVM for Linux VM guests?
> 
> I came across an old post comment yesterday (from
> http://echenh.blogspot.com/2010/04/how-to-extend-lvm-on-vmware-guest-os.html
> ) discussing the "hack" of LVM on Linux VM guests and whether it's
> better not to use it to simplify disk management.
> I've re-posted the comment below, does it sound reasonable? Is it
> better to not use LVM on Linux VM guests?
> 
> --Russell

I've had the same question.  I've decided to continue to use LVM, for these 2 
reasons:

1) my hypervisor, good, bad or indifferent, is VMware ESX 4.x and ESXi 4.x.  
Those hypervisors can't create virtual disks greater than 256 GB.  So, if I 
want a file-system larger than 256 GB, I have to have some other software - 
LVM, in this case.

2) I like being able to give disk devices descriptive names, like 
/dev/mapper/zimbra-data instead of simply '/dev/sdb' or similar.  There are 
probably ways other than LVM to do that, but LVM does offer that flexibility.

One thing I do avoid, however, is partitioning the virtual disks that might 
need to grow.  This is because of the pain described in part below.  The kernel 
often seems to have a hard time letting go of it's view of the partition table 
- either i have to umount the partition, or reboot.  However, if i use the disk 
unpartitioned, the kernel has no prob, and I can *extend and/or resize*fs 
without umount or reboot.

- Jon

> ---
> At my job, after doing the same kind of procedure graph, we began to
> ask ourselves, why are using a LVM on a Linux VM guests?
> 
> Since we're no longer living in the physical OS world, we didn't need
> to use the OS hacks(LVM) to overcome physical disk limitations
> anymore.
> We decided to Just let the hypervisor and virtual storage do that
> work for us.
> 
> For example, in our production setup (3 tier commerce with VMs for
> database , webserver, and appserver), we're see a great improvement
> in managability and performance (>10%) by just dropping LVM, and
> most partitions.
> 
> In your example, the resize process is 7 functional steps:
> 1. Increase size of VMDK
> 2. In VM OS, Create Partition (??)
> 3. REBOOT (!!)
> 4. PVCreate
> 5. VGExtend
> 6. LVExtend
> 7. Resize2fs
> 
> Going to a LVM/partition-less setup reduces expansion to 3 steps and
> we don't need to take the VM OS offline!
> 1. Increase size of VMDK
> 2- Inside the VM, OS, rescan the scsi drive with:'echo 1
> >/sys/class/scsi_device//rescan; dmesg' (dmesg will check that you
> drive isize has grown)
> 3- Resize2fs.
> 
> Our current disk arrangement has 3 VM HD devices
> 0 - small device (100M) with a single BOOT partition
> 1 - entire device is /
> 2 - entire device is SWAP
> 
> Doing this has simplified resizing so much, I now let the junior
> admins and my manager expand drive space as needed.
> 
> It's also let's us really be spartan on space since expansion is so
> quick. Instead of increasing systems in 30-50GB chunks, we can do
> 10-15GB and let our rmonitoring system warn us when space gets
> tight.
> -
> 
> ===
> Attention: The information contained in this message and/or
> attachments
> from AgResearch Limited is intended only for the persons or entities
> to which it is addressed and may contain confidential and/or
> privileged
> material. Any review, retransmission, dissemination or other use of,
> or
> taking of any action in reliance upon, this information by persons or
> entities other than the intended recipients is prohibited by
> AgResearch
> Limited. If you have received this message in error, please notify
> the
> sender immediately.
> ===
> ___
> CentOS mailing list
> CentOS@centos.org
> http://lists.centos.org/mailman/listinfo/centos
> 
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] is there a way to make the kernel see a new ethernet device without rebooting?

[Jon Detert]

- Original Message -
> From: "John Hodrien" 
> To: "CentOS mailing list" 
> Sent: Tuesday, September 27, 2011 9:08:09 AM
> Subject: Re: [CentOS] is there a way to make the kernel see a new ethernet 
> device without rebooting?
> 
> On Tue, 27 Sep 2011, Jon Detert wrote:
> 
> > I have a VMWare ESX server with virtual machines running CentOS.  I
> > want to
> > add an ethernet interface to one of the CentOS virtual machines.
> >  VMWare
> > allowed me to add a virtual NIC to the CentOS virtual machine while
> > it was
> > running.  However, the CentOS o.s. cannot see the new NIC.  Is
> > there a
> > command I can run to make the kernel aware of the new ethernet
> > device?  I
> > don't want to reboot the CentOS o.s. if I don't have to
> 
> Adding a NIC to VMWare with a CentOS 6 guest (e1000) works just fine
> here.  I


Adding the NIC in VMWare works fine for me too.  It's just that the guest
(CentOS) o.s. doesn't see the new NIC unless/until I reboot it.


> don't remember there being any problems with CentOS 5.
> 
> Does dmesg show evidence of the new device being attached?


No.

I wonder if it has to do with the type of NIC.  In my case, vmware says it's of
type 'flexible', and the CentOS o.s uses the 'pcnet32' driver for it.

- JD
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] is there a way to make the kernel see a new ethernet device without rebooting?

[Jon Detert]

I have a VMWare ESX server with virtual machines running CentOS.  I want to add 
an ethernet interface to one of the CentOS virtual machines.  VMWare allowed me 
to add a virtual NIC to the CentOS virtual machine while it was running.  
However, the CentOS o.s. cannot see the new NIC.  Is there a command I can run 
to make the kernel aware of the new ethernet device?  I don't want to reboot 
the CentOS o.s. if I don't have to

Thanks,

JD
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] proverbial floodgates thrown wide open for centos-announce list

[Jon Detert]

Starting maybe ~ 12 hours ago, the proverbial floodgates were thrown open on 
emails sent to the CentOS-announce mailing list for CentOS 5.x, all on account 
of individual package releases.

RedHat have not done the same.  What's up?

Are these mistakenly released re-announcements of (previously released) package 
updates?  Or, did the mailman server have a burp or hiccup?

Or, have the CentOS dev team been waiting for high tide before opening the 
flood gates?  If so, why?, and how about sending the mother-load of 
announcements as 1 email instead of 1/package, which would:

a) match the reality of the updates being released as a batch, and
b) keep my inbox from being flooded.  Yes, i could switch my mailman 
subscription to 'digest' mode, and probably will.  Still, I'm wondering why 
they were batch-released?

Regards,

Jon

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] selinux policy remnant according to /bin/ls on CentOS 6.0 box

[Jon Detert]

- Original Message -
> From: "James Edwards" 
> To: centos@centos.org
> Sent: Tuesday, September 20, 2011 12:52:34 PM
> Subject: Re: [CentOS] selinux policy remnant according to /bin/ls on CentOS 
> 6.0 box
> 
> On 9/20/2011 1:48 PM, Jon Detert wrote:
> > I installed CentOS 6.0 on 2 different x86_64 boxen.  Both
> > originally had selinux installed and enabled.  I never touched
> > selinux other than to remove as much of it as I could via rpm -e.
> >  As far as I can tell, here are the remaining packages that have
> > something to do with it:

-- snip --

> > However:
> >
> > 1) box1 still has files in /selinux whereas box2's /selinux is
> > empty;
> > 2) ls -l on box1 shows a '.' at the end of file/directory, which
> > means a SELinux security context applies, according to
> > https://fedoraproject.org/wiki/Fedora_11_FAQ#Why_does_ls_show_a_dot_.28..29_or_a_plus_.28.2B.29_at_the_end_on_the_file_modes_for_some_files.3F
> >
> > Any idea why box1 still seems to have an selinux policy applied,
> > and how to un-apply it?
> >
> > Thanks,
> >
> > Jon
> >
> Did you disable SELinux by changing 'SELINUX=disabled' in
> /etc/sysconfig/selinux?  Wouldn't that be easier than removing all

I did not do so explicitly.  But it is set to disabled as described above.
I assume the rpm -e did that.  So, there must be some other step missing.

As to that being easier: perhaps, had I known that file/setting existed.

> the
> RPMs?  If I may ask, is there a reason to removing the packages?

I do not plan to use them.

Less is more, right? 
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] selinux policy remnant according to /bin/ls on CentOS 6.0 box

[Jon Detert]

I installed CentOS 6.0 on 2 different x86_64 boxen.  Both originally had 
selinux installed and enabled.  I never touched selinux other than to remove as 
much of it as I could via rpm -e.  As far as I can tell, here are the remaining 
packages that have something to do with it:

# rpm -qa | grep -iE 'sel|pol'
checkpolicy-2.0.22-1.el6.x86_64
libselinux-2.0.94-2.el6.x86_64
libsepol-2.0.41-3.el6.x86_64
polkit-0.96-2.el6_0.1.x86_64
#

Both boxen have those packages.

However:

1) box1 still has files in /selinux whereas box2's /selinux is empty;
2) ls -l on box1 shows a '.' at the end of file/directory, which means a 
SELinux security context applies, according to 
https://fedoraproject.org/wiki/Fedora_11_FAQ#Why_does_ls_show_a_dot_.28..29_or_a_plus_.28.2B.29_at_the_end_on_the_file_modes_for_some_files.3F
 

Any idea why box1 still seems to have an selinux policy applied, and how to 
un-apply it?

Thanks,

Jon
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] questions about upgrading

[Jon Detert]

I'm pretty new to the task of upgrading CentOS or RedHat o.s.'s.  I'm more 
familiar with upgrading Debian based o.s.'s.  Here are my questions:

1) will yum ever upgrade the major version?  I understand that there is no 
upgrade path from v5.x to v6.x, but is that always the case?  e.g. was there an 
upgrade path from v4.x -> v5.x?  If yum is capable of doing so, what are the 
command-line arguments necessary?

2) can I upgrade a minor version to any other greater minor number, or only to 
the greatest?  E.g. if i have a v5.4 box, can i upgrade to v5.5 or v5.6, or 
only to v5.7 (currently the latest)?

3) if an upgrade delivers a new kernel, am I obligated to run it (i.e. will the 
old kernel remain installed, or will the upgrade remove the old kernel)?  If 
not, does the upgrade automatically update grub such that the new kernel 
becomes the default?  Will the upgraded box function properly without running 
the new kernel?

4) aside from when an upgrade includes a new kernel, how can I tell when a 
reboot is necessary?  I gather that a new glibc may require services to be 
restarted, but is a reboot strictly necessary?

5) is there a way to apply only the updates that were made for the given minor 
version you're currently at?  E.g. if I'm at v5.5, how can I only apply updates 
that were made prior to v5.6's availability?

Thanks,

Jon
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] trouble building an rpm

[Jon Detert]

- Original Message -
> From: "Andrew Dorozhkin" 
> To: "CentOS mailing list" 
> Sent: Monday, September 12, 2011 11:39:44 AM
> Subject: Re: [CentOS] trouble building an rpm
> 
> 12.09.2011 20:00, Jon Detert wrote:
> > i'm a newbie at building rpms.  I made a few rpms years ago, but
> > can't now make one on Centos 5.
> >
> > In /usr/local/rpmbuild, there are 5 subdirs: BUILD  RPMS  SOURCES
> >  SPECS  SRPMS  tmp
> >
> > When I type: rpmbuild -ba
> > /usr/local/rpmbuild/SPECS/centos-release-cr-ihc.spec
> >
> > I get 'file not found' errors, for paths in
> > /usr/local/rpmbuild/tmp:
> > error: File not found:


-- snip --

> > my spec file is
> > /usr/local/rpmbuild/SPECS/centos-release-cr-ihc.spec
> > the contents of which are:
> > Summary: Package to set up IHC use of the centos-cr repository

-- snip --

> > %build
> > %install
> > install -m 0755 -d /etc/yum.repos.d
> > install -m 0755 etc/yum.repos.d/CentOS-cr.repo

-- snip --

> Directory
> /etc/yum.repos.d
> and file
> /etc/yum.repos.d/CentOS-cr.repo
> need to be installed into $RPM_BUILD_ROOT and not into actual  /
> during
> the %install phase.
> 
> Try something like this:
> 
> %install
> install -m 0755 -d $RPM_BUILD_ROOT/etc/yum.repos.d
> install -m 0755 etc/yum.repos.d/CentOS-cr.repo
> $RPM_BUILD_ROOT/etc/yum.repos.d

Thanks!  That was the problem, and your solution worked.

Regards,

Jon
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] trouble building an rpm

[Jon Detert]

i'm a newbie at building rpms.  I made a few rpms years ago, but can't now make 
one on Centos 5.

In /usr/local/rpmbuild, there are 5 subdirs: BUILD  RPMS  SOURCES  SPECS  SRPMS 
 tmp

When I type: rpmbuild -ba /usr/local/rpmbuild/SPECS/centos-release-cr-ihc.spec

I get 'file not found' errors, for paths in /usr/local/rpmbuild/tmp:
error: File not found: 
/usr/local/rpmbuild/tmp/centos-release-cr-ihc-buildroot/etc/yum.repos.d
error: File not found: 
/usr/local/rpmbuild/tmp/centos-release-cr-ihc-buildroot/etc/yum.repos.d/CentOS-cr.repo
RPM build errors:
File not found: 
/usr/local/rpmbuild/tmp/centos-release-cr-ihc-buildroot/etc/yum.repos.d
File not found: 
/usr/local/rpmbuild/tmp/centos-release-cr-ihc-buildroot/etc/yum.repos.d/CentOS-cr.repo

All i did was to:
1) put a tarball in SOURCES,
2) put a spec file in SPECS,
3) rpmbuild -ba my spec file

Did I miss a step?  If not, here's the contents of my tarball and spec file:

The tarball in SOURCES is named: centos-release-cr-ihc-1.tar.gz
the contents of which are:
drwxr-xr-x root/root 0 2011-09-09 15:40:01 centos-release-cr-ihc-1/
drwxr-xr-x root/root 0 2011-09-09 15:39:57 centos-release-cr-ihc-1/etc/
drwxr-xr-x root/root 0 2011-09-09 15:40:39 
centos-release-cr-ihc-1/etc/yum.repos.d/
-rw-r--r-- root/root   523 2011-09-09 15:35:59 
centos-release-cr-ihc-1/etc/yum.repos.d/CentOS-cr.repo

my spec file is /usr/local/rpmbuild/SPECS/centos-release-cr-ihc.spec
the contents of which are:
Summary: Package to set up IHC use of the centos-cr repository
Name: centos-release-cr-ihc
Version: 1
Release: 1
Source0: centos-release-cr-ihc-1.tar.gz
License: GPL
Group: IHC-config
BuildArch: noarch
BuildRoot: %{_tmppath}/%{name}-buildroot
%description
Package to set up IHC use of the centos-cr repository
%prep
%setup -q
%build
%install
install -m 0755 -d /etc/yum.repos.d
install -m 0755 etc/yum.repos.d/CentOS-cr.repo /etc/yum.repos.d/CentOS-cr.repo
%clean
rm -rf $RPM_BUILD_ROOT
%post
echo " "
echo "Yay IHC IS Team!"
%files
%dir /etc/yum.repos.d
/etc/yum.repos.d/CentOS-cr.repo

Thanks,

Jon
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Grep: show me this line and the next N lines?

[Jon Detert]

- Original Message -
From: "Dotan Cohen" 
To: "CentOS mailing list" 
Sent: Tuesday, June 7, 2011 1:20:20 AM
Subject: Re: [CentOS] Grep: show me this line and the next N lines?

On Tue, Jun 7, 2011 at 05:26, Kenneth Porter  wrote:
> --On Tuesday, May 31, 2011 1:08 AM +0300 Dotan Cohen 
> wrote:
>
>> Can grep show the matching lines and the next N lines after a match?

grep -A N 

see http://linux.die.net/man/1/grep
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] problems creating read-only, 'consumer' dirsrv replica

[Jon Detert]

Hello,

I am trying to deploy an additional read-only replica (aka.
'consumer') in a single-master dirsrv environment.

The master, and the other pre-existing consumer servers, are all
'fedora-ds' running on Fedora 7.  I'm trying to add a consumer running
on Centos 5.5.  Ultimately, I intend to replace the Fedora ds servers
with Centos dirsrv servers.

I'm trying to deploy the Centos dirsrv consumer without running the
dirsrv-admin.

I've been following the directions at
http://docs.redhat.com/docs/en-US/Red_Hat_Directory_Server/8.1/html/Administration_Guide/index.html
, specifically in chapter 8.  When I get to "8.7.2. Configuring
Consumers from the Command Line" (
http://docs.redhat.com/docs/en-US/Red_Hat_Directory_Server/8.1/html/Administration_Guide/Managing_Replication-Configuring-Replication-cmd.html#Configuring-Replication-Consumers-cmd
),  I run into problems:

1) ldapmodify complains that I'm missing the required attribute
'nsds5replicaId'.  So I add it, and guess at the replicaID number (cuz
the guide says nothing about the ID).

2) even after resolving issue 1) above, I get an error: "ldap_add: No
such object"
The return value of the ldapmodify command is 32.
The changetype of the operation is 'add', so ldapmodify knows it's
trying to add an object.

The input I pass to ldapmodify is just like shown in the guide, but w.
the addition of the nsds5replicaId attribute mentioned in issue 1)
above:
dn: cn=replica,cn="dc=infinityhealthcare,dc=com",cn=mapping tree,cn=config
changetype: add
objectclass: top
objectclass: nsds5replica
objectclass: extensibleObject
cn: replica
nsds5replicaroot: dc=infinityhealthcare,dc=com
nsds5replicatype: 2
nsDS5ReplicaId: 65535
nsds5ReplicaBindDN: uid=RManager,cn=config
nsds5flags: 0

Any idea what's wrong?  Do I have to first create dn:
cn="dc=infinityhealthcare,dc=com",cn=mapping tree,cn=config ?  If so,
how?  The guide doesn't seem to have a step for that, except for when
adding a 'hub' dirsrv.  However, I mean to add a 'consumer', not a
'hub'.

Thanks,

Jon
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos