19 January 2014 – Sunday


  • Returned home from a much needed weekend away to find that two storage nodes (stornode3 and stornode4) were hung in the Linux NFS client code (in the kernel of course).  Stornode3 successfully power cycled from remote.  However, stornode4 had a POST error that required physical presence at the console to go into setup and verify that the default settings that had been chosen (apparently it failed to read the nvram???) were ok before it would boot.  Sadly, that function is not available via the IMM, which is yet another annoyance with the new systems — they are not completely remotely manageable even though they claim to be.
  • Both stornode3 and stornode4 are now back at 64K block sizes for wsize/rsize on the NFS client mount parameters in /etc/fstab.


Leave a Reply

You must be logged in to post a comment.