Author Archives: chouse

VCAP5-DCA exam experience

I sat the VCAP5-DCA exam yesterday, also known as VDCA510. Actually taking the exam fulfills a journey I started on back in 2011 when I first registered for the vSphere 4 version of the exam. However, taking a new and busy job, plus life in general caused me to cancel and ultimately schedule the vSphere 5 version of the exam, and then reschedule it a half-dozen times.

By the time it finally rolled around, I was sick of studying and thinking about it and ready to just get it done and over with.

It was a good experience and I’m pretty sure I passed, and hopefully did pretty well. I tried the tactic of going through all the questions and writing down briefly what topic they were about, whether they were “administrative”, “configuration”, or “troubleshoot”. That was helpful to get an idea of what would be coming up as I moved through the exam, but ultimately I just started at question 1 and worked my way through the rest of them.

I was grateful to have the vSphere 5 PDFs available with Adobe Reader to search through them and am not ashamed to admit that I did refer to a few things in there, especially when it came to some PowerCLI stuff.

I believe I answered every question correctly and accurately, and only ended up skipping one due to time. I had circled it to come back and deal with later as it was not integral to future questions, but with 5 seconds remaining, I went ahead and clicked Finish, so oh well. I did managed to complete 2 that I had circled to return to, so that was satisfying.

The Pearson Vue site had nice 20-inch widescreen monitors, but unfortunately the PC or remote desktop was only configured to run in 4:3 so the resolution was something like 1280×1024 and could not take advantage of the widescreen monitor. This was annoying when I had many different windows up and was bouncing between things. My advice, minimize everything as soon as you’re done so you can easily call it back up from the rows of minimized windows at the bottom (no start bar). Much better than hunting through a bunch of open windows to find the one you need.

One other issue was that during one of the questions, I had to press a function key within a VM console, but none of the function keys seemed to work – as if Pearson or something had disabled them. Using F2 to try to rename a file in a remote desktop did not work either. I emailed the Certification team to share my experience so hopefully I will still get full credit for the question.

The resources that I relied on were:

Probably most helpful was my day to day vSphere administration that I do for various customers. However, as with any exam, there were questions on areas that I am just not too familiar with because my customers don’t use those pieces, so I did have to lab them up and play with them a bit.

What I liked about the exam was that I was free to accomplish the question in whichever manner I was most comfortable with, which was via the GUI. But some questions were specific to using the CLI, so those had to be done via that, and I was glad to have the documentation to refer to.

On to the VCAP5-DCD!

esxupdate error code 15 – copying payloads to altbootbank

Recently had a Cisco UCS blade running vSphere 5.0 Update 2 fail to install a patch via Update Manager. The other hosts in the cluster installed the patch fine and rebooted safely, but this one did not.

Upon review of the host’s /var/log/esxupdate.log, we find errors during the patch installation:

2013-07-16T13:14:36Z esxupdate: BootBankInstaller.pyc: INFO: Copying /bootbank/scsi-qla.v00 to /altbootbank/scsi-qla.v00
2013-07-16T13:14:36Z esxupdate: esxupdate: ERROR: An esxupdate error exception was caught:
2013-07-16T13:14:36Z esxupdate: esxupdate: ERROR: Traceback (most recent call last):
2013-07-16T13:14:36Z esxupdate: esxupdate: ERROR:   File "/usr/sbin/esxupdate", line 216, in main
2013-07-16T13:14:36Z esxupdate: esxupdate: ERROR:     cmd.Run()
...
2013-07-16T13:14:36Z esxupdate: esxupdate: ERROR: InstallationError: ("set(['VMware_bootbank_net-nx-nic_4.0.557-3vmw.500.1.11.623860', 'VMware_bootbank_scsi-rste_2.0.2.0088-1vmw.500.1.11.623860', 'VMware_bootbank_scsi-megaraid-mbox_2.20.5.1-6vmw.500.1.11.623860', 'VMware_bootbank_scsi-mptsas_4.23.01.00-5vmw.500.1.18.768111', 'VMware_bootbank_block-cciss_3.6.14-10vmw.500.1.11.623860', 'VMware_bootbank_ipmi-ipmi-si-drv_39.1-4vmw.500.2.26.914586', 'VMware_bootbank_ata-pata-amd_0.3.10-3vmw.500.1.11.623860', 'Broadcom_bootbank_misc-cnic-register_1.72.1.v50.1-1OEM.500.0.0.472560', 'VMware_bootbank_ipmi-ipmi-msghandler_39.1-4vmw.500.1.11.623860', 'VMware_bootbank_esx-tboot_5.0.0-2.26.914586', 'VMware_bootbank_scsi-mpt2sas_06.00.00.00-6vmw.500.1.11.623860', 'VMware_bootbank_uhci-usb-uhci_1.0-3vmw.500.1.11.623860', 'VMware_bootbank_scsi-mptspi_4.23.01.00-5vmw.500.1.11.623860', 'VMware_bootbank_net-r8168_8.013.00-3vmw.500.1.11.623860', 'VMware_bootbank_ohci-usb-ohci_1.0-3vmw.500.1.11.623860', 'VMware_bootbank_ima-qla4xxx_2.01.07-1vmw.5
...
2013-07-16T13:14:41Z esxupdate: HostImage: DEBUG: Live image has been updated but /altbootbank image has not.  This means a reboot is not safe.

Ignoring the “a reboot is not safe” message, I bounced the host hoping that a reboot would clear up whatever problem existed and the patch would then install successfully. Nope!

boot fail after update

This lovely purple boot error screen was displayed right after “VMkernel started successfully.” No amount of ALT+F12 etc would work. Ultimately I opened a case with VMware and per an internal KB article, this can occur with Cisco UCS systems booting from FC SAN, as was happening in this case:

Symptoms
A freshly installed ESXi 5.0 host does not load after a reboot.
If you are booting from FC SAN and having hardware issue the ESXi 5 host can purple screen after the reboot.

Note: This issue is seen to occur during the Cisco UCS system boot from FC SAN scenario and the system file gets corrupted.

The ESXi console shows a purple screen that says:

The system has found a problem on your machine and cannot continue. Could not populate the filesystem: Already exists

Cause
This issue may be caused by a BIOS remap or a glitch of the Raid Array BIOS.

Resolution
To resolve this issue, perform an upgrade install on the ESXi host:

Boot on ESXi 5 install media.
Choose upgrade ESXi and preserve the vmfs datastore.

The host is rebuilt according to the new hardware/BIOS layout.

Additional Information
If the host fails with a purple screen when booting from FC SAN, you may have to perform an upgrade install to repair the installation.

I rebooted the blade and booted off the Cisco custom ESXi installer media, even though the build was quite old. I performed an Upgrade on the existing ESXi install on its boot LUN and after it was done, the blade booted successfully. Its configuration was intact and its build number was current.

I tried installing the patch again but it also failed, but at least with a different error message:

esxupdate: esxupdate: ERROR: InstallationError: ('', 'There was an error checking file system on altbootbank, please see log for detail.')

According to VMware KB 2033564, running dosfsck -a -w against the altbootbank partition will fix any errors and then the update can be attempted again.

To find the altbootbank disk, follow the instructions in the KB article or look in the esxupdate.log:

2013-07-16T13:11:12Z esxupdate: vmware.runcommand: INFO: runcommand called with: args = '['/sbin/dosfsck', '-a', '-v', '/dev/disks/naa.60060e8006d022000000d0220000110c:6']', outfile = 'None', returnoutput = 'True', timeout = '0.0'.

The “naa.600…” is the altbootbank – don’t forget the :6 (partition number) at the end!

Once dosfsck comes back clean, attempt the update install again and it should now be successful, including a successful reboot.

Answering VM questions with PowerCLI

Managing Virtual Machine questions with PowerCLI | VMware PowerCLI Blog – VMware Blogs.

Seems like every day there are a number of View VMs that have the question asking if the VM was moved or copied. VMware View can’t do any operations with them until the question is answered, and it won’t answer the question itself so here’s a quick one-liner to just answer them all at once:

Get-VMQuestion | Set-VMQuestion -Option "I copied it" -confirm:$false

This could be scripted to run regularly, but for now I only run it when I see a few pending questions.