[Libguestfs] P2Vs seem to require a very robust Ethernet

Greg Scott GregScott at Infrasupport.com
Sat Dec 3 04:32:34 UTC 2011


Here's another nugget of info.  Watching top on my conversion server, I
consistently see roughly 50% id and 50% wa.  It varies second by second
by a percent or two but it's generally around 50-50.  The "wa" state
means waiting for an I/O to complete so it seems like something is
taking a **very** long time to complete IOs.  My guess is, something is
stalling virt-p2v-server.  I wonder if there's a way to drill down into
an individual process the way we used to do it with OpenVMS?

Every once-in-a-while, virt-p2v-server will jump up to the top of the
top list.  The current P2V has been running for about 6 hours so far and
I don't know what the state looks like on the source server console.
The logfile on the conversion server is still blank.  Watching the
Export domain with my RHEV Manager, I see this P2V has used about 28 GB
of space on the NFS server so far.

On the source server - how do we know the NIC speed and duplex?  I guess
we can look at the switch port from the switch's point of view, but how
would we look at it and/or set it from the host point of view?

- Greg




More information about the Libguestfs mailing list