This is a presentation by Nick Weaver at PuppetConf 2013, although it really doesn’t have to much to do with Puppet I found it to be an interesting talk on the whats and whys of automation.

At the end of the presentation Nick gives his recommendations on how to get started in the field, my 2 cents would be to learn two languages. Learning one is a great start but I think a second will really give you extra flexibility. Most vendors seem to support PowerShell (not always very well) and Ruby. Honestly if you are a pure Windows shop (which is exactly 0 people) then just stick with PowerShell, but if you need need to venture off into supporting anything else then another language would really help with adaptability, just so long as it gets the job done.

I seem to be on a bit of a Puppet kick as of late. Of a night I’ve been watching the many awesome videos of the 2013 PuppetConf and I found this one quite interesting. To me this pattern shows the future of deploying vendor supplied applications and I’m sure all sysadmins will be eternally grateful.

I just wish the presentation had been allowed to finish. :(

For some time now I’ve been designing and developing some labs to replace the coming void that will be left by TechNet (please sign the petition!). My efforts have been focused on using VMware Fusion in conjunction with Vagrant and Packer. Both Vagrant and Packer are big deals in the DevOps communities due to the ability to quickly and easily destroy and create lab/dev environments.

Due to the rapid rebuilds that can occur heavy automation is a must, and this is where Vagrant really shines. Vagrant is essentially a controller for you virtual lab and Packer is a tool that can build virtual machine templates in an automated fashion.

If you work in an environment with lots of virtualised branches then at some point you are going to want to add VMware Update Manager (VUM) baselines to all or most of them. This is pretty easy to do with PowerCLI and the VUM snapin. If you want your baseline to apply to everything then I’d recommend you just attach the baseline to the vCenter object at the top of the tree. But I need to exclude one data centre and thus this code:

• If you are running a Dell EqualLogic SAN and are on an effected version then you need to lookout for this VMFS Heartbeat corruption issue with vSphere 5.5. The fix is to upgrade your SAN firmware.

• It appears there is also an issue with EMC VNXe SANs and iSCSI. This one only effects dual SP systems. I don’t have a direct link for this one but the EMC KB is 169041, Cormac Hogan has a few more details on his blog. There is a hotfix for this issue.