Solved

VM Migration fails with "Reader is unavailable" error

  • 27 February 2018
  • 3 replies
  • 1574 views

Badge +4
Xtract 1.1.2. Was migrating a VM (Windows 2008R2, MDT server with large second drive for WSUS patching) from ESXi 6.0 to AHV 5.1.4 cluster with Xtract deployed on AHV cluster. The migration ran for about 3.5 hours and seeded 240GiB of the VMs 580.5GiB size. (mentioning drive size as unsure of it as potential focal point) I tried refreshing the Source and Target environments to ensure Xtract still has connectivity, no issues. Retrying the migration for the VM results in the same error "Reader is unavailable". Only thing outside of completely allowing Xtract to migrate the VM is I manually removed VMware tools and Installed VirtIO drivers, so I didn't provide credentials for Xtract to do this part as I'm looking to just move the data... Had initially tried vCenter to export VM as OVF but that appears to have flopped because I can't convert the vmdk with qemu and it's size was much smaller than anticipated (thus I think it failed to fully export).

Posting this as I opened a Support Case to review the error and because this may assist others as I expect use of Xtract to increase.

After starting to write this thread i saw there was a new version (today!) so I upgraded to 1.1.3. I decided to reboot the VM as well. The upgrade was simple, I decided to log in to the Xtract VM with the default credentials in the Admin Guide and reboot the VM. I couldn't connect to the web GUI, so did an ifconfig and didn't see my inet (IPv4) address, so rebooted again and it showed up. Connected back into the web GUI, restarted the migration and noticed a nice "new" thing with a hour(s) minute(s) remaining along with the percentage progress indicator! So it's back running now with roughly 5 hours to go (moving over 1Gbps network).

TL;DR - "Reader is unavailable" error, try restarting the Xtract VM (ensure your jobs are paused).
icon

Best answer by BG_selfcloud 23 March 2018, 14:35

I did. The final response after reviewing logs was:



"From what I am seeing in the logs it appears that the Xtract diskreader lost connection with the source and was not able to reconnect.



This same signature (shown below) has been seen in other cases caused by duplicate IPs, network disruptions, and other network related issue."



So I chocked it up to some network issue and rebooting the Xtract appliance allowed the pathing to fully re-establish.
View original

This topic has been closed for comments

3 replies

Userlevel 2
Badge +5
Hi BG_selfcloud - did Nutanix Support help you get this issue solved?

I don't know if the manual removal of VMware tools played a part in the overall issue, however I suspect it would have impacted the migration duration without the correct more-performant network device drivers.

I will checkin with PM to see if they have any suggestions, however please do open a Support ticket if you haven't already, as they are best placed to provide the necessary support.

-thanks!
Badge +4
I did. The final response after reviewing logs was:

"From what I am seeing in the logs it appears that the Xtract diskreader lost connection with the source and was not able to reconnect.

This same signature (shown below) has been seen in other cases caused by duplicate IPs, network disruptions, and other network related issue."

So I chocked it up to some network issue and rebooting the Xtract appliance allowed the pathing to fully re-establish.
Userlevel 2
Badge +5
Thanks @BG_selfcloud - hopefully a one-off issue that doesn't rematerialise, but let us know if it does!