Saturday, March 8, 2014

Matahari is deprecated

After a nice long hiatus... Matahari is deprecated. 

I was spinning up a VM to play with and happened to have a CentOS 6.2 ISO to do the install from. Once it was complete, I went to update the OS and received the following errors:

 ---> Package qpid-cpp-server-ssl.x86_64 0:0.12-6.el6 will be updated
---> Package qpid-cpp-server-ssl.x86_64 0:0.14-22.el6_3 will be an update
---> Package qpid-qmf.x86_64 0:0.12-6.el6 will be updated
---> Package qpid-qmf.x86_64 0:0.14-14.el6_3 will be an update
--> Finished Dependency Resolution
Error: Package: matahari-host-0.4.4-11.el6.x86_64 (@anaconda-CentOS-201112091719.x86_64/6.2)
           Removing: qpid-cpp-client-0.12-6.el6.x86_64 (@anaconda-CentOS-201112091719.x86_64/6.2)
           Updated By: qpid-cpp-client-0.14-22.el6_3.x86_64 (base)
               Not found
Error: Package: matahari-host-0.4.4-11.el6.x86_64 (@anaconda-CentOS-201112091719.x86_64/6.2)
           Removing: qpid-cpp-client-0.12-6.el6.x86_64 (@anaconda-CentOS-201112091719.x86_64/6.2)
           Updated By: qpid-cpp-client-0.14-22.el6_3.x86_64 (base)... 
           Removing: qpid-cpp-client-0.12-6.el6.x86_64 (@anaconda-CentOS-201112091719.x86_64/6.2)

           Updated By: qpid-cpp-client-0.14-22.el6_3.x86_64 (base)
               Not found

 You could try using --skip-broken to work around the problem

 You could try running: rpm -Va --nofiles --nodigest

[root@centosx64-py3 ~]#

So, I did some digging and eventually stumbled across a thread with a note that "Matahari is deprecated and the upstream vendor says you should remove it:"
 # yum remove matahari*
I removed matahari, ran the yum update again, and it worked like a charm. 

Sunday, February 3, 2013

New position takes 120% of my bandwidth

So, I started a new position as a Linux Admin back in August of last year. The company is a fortune 100 and the job is fantastic. The only down side is it literally consumes more time than I have available. By this I mean we are currently working on such a wide spectrum of technologies that it's hard to give attention to anything else.

That spectrum I mentioned spans from  Openstack / vSphere Admin'ing to international remote site infrastructure installations. There's a lot to do. Which keeps me busy and happy. The single caveat is that my available attention to things like certs has been greatly diminished. I had to drop last terms Sec+ and VCP classes as I just couldn't keep up with the ramp up from the new job and class work.

Needless to say the blog was set to the side and only lately have I begun to surface. However my current focus is spread a bit thin. I'm looking over widely distributed directory deployments, working on the Python Scripting Expert class, and reading Pro Puppet.

Things seem to move very fast in the company and if I'm not involved and actively working on the technology of interest the project moves on without me. Puppet is a good example of this. We are in a slow migration from CF Engine to Puppet, but if I don't jump in on configs, then the task will get picked up by another Admin. I loose out on the experience and knowledge. On the other side, I've got a full plate with Perl scripts, vSphere work, PM work for a Nagios deployment, Nagios reporting research, puppet manifests to write, plus more. Never a dull moment!

So, for the next while I'll be doing my best to drop my usual random notes and articles, but please bare with me. The timing may be more elongated between posts as I'm running fairly short on that these days.

Be well, grab your towel, and don't panic!

Saturday, September 22, 2012

Sleeping Lions

I've had a Mac Pro for about 4 years now and have converted my house to an almost entirely Linux/Mac environment. The only exception is the Win VM that is needed for the Public library's audio books. Wheat I've noticed with OS X is that with each upgrade Apple likes to change little things for no reason what so ever. It's an odd practice and one that causes minor head aches each feline upgrade.

The example is the sleep mechanism in OS X. 10.5 was stable, 10.6 stable still, 10.7 annoying change, 10.8 random settings change.  Why, Apple, why!? At any rate, here's an older post I was going to publish in 2011 that went unfinished and unpublished, but was the reference I needed to fix a new 10.8 issue.

Begin 2011 post:

My main workstation is a Mac Pro 2008. Its the dual quad core intel with 14gb of ram. I run OS X and a Windows XP VM / CentOS 5.5 VM on it at all times. So, when Lion (OS X 10.7) came out the other day I was interested and picked it up via the App store download. I did the upgrade, all went fine. Well, that is except for sleeping.

I usually put the machine to sleep each evening when I retire to bed. It's worked that way since I bought the machine via the small business rep at the Apple store. I like it and am used to it. Well, this didn't work in OS X 10.7. The monitor would sleep, but not the machine itself. So, I did some googleing and eventually came across a post detailing using the pmset command to see what the power management settings were. It looked something like this:

Kim-mp:etc kim$ pmset -g
Active Profiles:
AC Power        -1*
Currently in use:
 hibernatemode    0
 disksleep    10
 womp        0
 networkoversleep    0
 sleep        0 (imposed by 628)
 powerbutton    1
 ttyskeepawake    1
 hibernatefile    /var/vm/sleepimage
 autorestart    0
 panicrestart    157680000
 displaysleep    60

Curious, that process ID#628 is stopping the sleep process from taking place. Ok, what's 628?

Kim-mp:~ kim$ ps -ef | grep 628
    0   628     1   0 10:44PM ??         1:56.59 /usr/sbin/smbd

Huh. Samba is holding up my shutdown? What the? Ok, what's up with Lion and Samba now? I've got a windows based NAS (I know, its headless though and my freenas hasn't gotten set up yet) and my Windows VM that share via SMB. Jeez.

So,  a little more googling lead to an Anandtech review of OS X Lion (here!) that gave good insight into what has happened and why. Ok, no more Samba on Lion for a while till the kinks are worked out. I read on Apple's community support forums the bugs should be gone by 10.7.3 (oh joy!). So, it's either AFP or NFS. Ok, really it's just NFS so my Linux machines would be served as well.

//End 2011 post

Well, I upgraded to Mountain Lion the other day and wouldn't you know it, sleep issue again. I would put the machine to sleep, go to work or bed and when I returned to the workstation, it would be on. My first thought was the dang cat was sitting on my KB again (he and I have been going through some lessons on things he can do and things he can't).  However, today I put the machine into sleep mode and the cat was in one of the bedrooms for the majority of the day. Ok, I need to dig deeper.

So, I looked up the 2011 post above and ran the pmset -g command again.
Kim-mp:tmp kim$ pmset -g
Active Profiles:
AC Power        -1*
Currently in use:
 hibernatemode        0
 womp                 1
 networkoversleep     0
 sleep                0
 powerbutton          1
 ttyskeepawake        1
 hibernatefile        /var/vm/sleepimage
 autorestart          0
 disksleep            10
 displaysleep         60
No process issues like before. So, I gave a look over the settings.  No more panicrestart. Interesting. Womp is set to 1, but in 2011 it was set to 0. A quick man reference showed that womp was WOL during sleep. Ahhh, set to 0. Everything else looked good.

Kim-mp:tmp kim$ sudo pmset womp 0

WARNING: Improper use of the sudo command could lead to data loss
or the deletion of important system files. Please double-check your
typing when using sudo. Type "man sudo" for more information.

To proceed, enter your password, or type Ctrl-C to abort.

Kim-mp:tmp kim$ pmset -g
Active Profiles:
AC Power        -1*
Currently in use:
 hibernatemode        0
 womp                 0
 networkoversleep     0
 sleep                0
 powerbutton          1
 ttyskeepawake        1
 hibernatefile        /var/vm/sleepimage
 autorestart          0
 disksleep            10
 displaysleep         60
Kim-mp:tmp kim$

All should be good now and hopefully Mountain Lion will finally stay asleep.

Update: It verks!

Saturday, September 8, 2012

VMware Certification Roadmap And A Quick Update

I've been at the new position for about a month now. I'm one of the Admins for a global Linux based network. There's no shortage of oddities to deal with around the globe. I'm enjoying it and it's really nice to be with like minded folks (nerds!). I've had to adjust to a larger environment moving from a 2 man team to a 6-7 man team in a rather large company. It's going to be fun.

I've started some classes as well. A second Network Security course which is very Windows centric. I suppose when you're trying to patch holes you go for the most breakable OS.  I plan to sit for the Security+ exam after this class and will hopefully end up with a voucher from the course. Of course this means I'll be giving some more money to Darril Gibson for his audio material and ebook covering the SY-301. This will finish up my security courses for a while until I start work on the CCNA:Security sometime in the next year or two.

My second class is the VMware vSphere: Install, Configure, and Manage V5. It's being offered at my local community college for a fraction of the normal 1-3K cost. It does however come with the caveat of being 16 weeks long. You do most of the labbing in a "Pod" which is a virtual setup of a small vSphere installation. It's a not huge difference from deploying Openstack locally and having a private cloud at your disposal. As I understand it though, the VCP5 exam (VMware Certified Professional) requires a good amount of studying in order to pass the exam. That said, it's not just the VCP anymore.

Thc VCP exam has been broken out into 4 different flavors (and I might be late to this game), VCP IaaS, VCP Cloud, and VCP desktop, and VCP datacenter virtualization. 

VMware's new certification roadmap:

and of course, links to all the exams:

Thursday, August 9, 2012

Annual hiatus + new job = lack of posts.

Hello World. Hope all is going well.

I've been quite busy as of late. In June I happened upon an opportunity to move to a larger, more fast paced company. I was all set to move companies. Then in mid July I was extended an offer with a rather large fortune 100 to work as a Unix Admin and couldn't pass up the offer. My last few month (and possibly next few) have been consumed with documenting my old position, taking a brief break, and working on my new position. It's all been keeping me very busy.

Classes start at the end of August as well. This includes VMWare's Install, Configure, and Manage, a Security+ course, and a required Nutrition course. I have an extremely full plate as my new position also includes a bit of international travel. What will be nice though is the study time I'll be able to get in while abroad. Still though, a full plate.

My posts will be hit and miss as I pick up the new environment. Though, I hope to pick up again as I get settled.

Friday, June 22, 2012

Nagios issues, duh.

I went to go pull up the Nagios web interface yesterday and to my surprise received a "Error: Could not read object configuration data!". Troubleshooting here we come!

First things first, check the process
 # /etc/init.d/nagios status  
 nagios is stopped  

Not running? Odd. ok, start it up
 # /etc/init.d/nagios start  
 nagios is stopped  
 Configuration validation failed              [FAILED]  

 # /usr/bin/nagios -v /etc/nagios/nagios.cfg  
  Total Warnings: 0  
 Total Errors:  0  
 Things look okay - No serious problems were detected during the pre-flight check  

So, I checked the log files and they are all from past dates, including nothing from the last day or so.
 # cat /var/log/nagios/nagios.log | perl -pe 's/(\d+)/localtime($1)/e' | less  

Nifty perl script there to fix the dates, but it's not mine, just a web snippet.

 I was stumped. Then I had a thought, SELinux errors...
 # getenforce  

 # setenforce permissive 

and give it another shot,
 # /etc/init.d/nagios start  
 nagios is stopped  
 Starting nagios:                      [ OK ]  

As they say on Afrotechmods,

A lot of folks disable SELinux in their environments as it can get rather complicated. However, for this issue I was able to dig up this post regarding setting up a new type definition for Nagios, but haven't yet tried it myself.

Monday, June 18, 2012

Python and My Summer Schedule

With the RHCSA finished up I am digging in no Python before moving on to the RHCE. Plus, I have a few months before my VMWare ICM (v.5) class, which begins at the end of August.

The first class I'm taking is Zed Shaw's Learn Python The Hard Way. It is essentially a book containing some 50+ exercises for learning Python. I came across the course on Udemy back around the end of 2011 and haven't had a chance to start the class until now. The Udemy course has Q&As with Shaw and videos detailing the exercises in the book. The course starts out very basic using Python 2.x and covering such things as printing text and doing mathematics. Also, the course is setup to be a "learn by doing" course with many items not explained. You are, however, instructed to look these items up and Mr. Shaw seems ready to anser questions at a moments notice. I've given myself about a month to blow through the course and do all the exercises a few times.

SecurityTube Python Scripting Expert from Vivek Ramachandran on Vimeo.

In July I will begin the Securitytube Python Scripting Expert (Spse) Course And Certification which seems to jump into more in depth Python topics with an infosec focus. That should take me through the end of July when I'm tentatively planning to take the SPSE exam.

After that August will have arrived. I'll begin studying for the RHCE exam then and come late August my VMware ICM class will begin. The ICM course is one night, weekly, but I have 2 other courses as well. So,  hopefully I will have some additional time to study for the RHCE durring the initial few months of classes.