On 9/19/19 4:50 AM, Richard W.M. Jones wrote:
On Wed, Sep 18, 2019 at 01:59:01PM +0100, Richard W.M. Jones wrote:
> We have a running problem with the nbdkit VDDK plugin where the VDDK
> side apparently disconnects or the network connection is interrupted.
> During a virt-v2v conversion this causes the entire operation to fail,
> and since v2v conversions take many hours that's not a happy outcome.
>
> (Aside: I should say that we see many cases where it's claimed that
> the connection was dropped, but often when we examine them in detail
> the cause is something else. But it seems like this disconnection
> thing does happen sometimes.)
It turns out in the customer case that led us to talk about this, a
Checkpoint firewall was forcing the VDDK control connection to be
closed after an idle period.
Would TCP keepalive help here?
(The VDDK connection as a whole was not
actually idle because data was being copied over the separate data
port, but the firewall did not associate the two ports). I believe
nbdkit-retry-filter would have helped in this case because reopening
the VDDK connection will reestablish the control/metadata connection,
and therefore I am looking at an implementation now.
Compared to re-establishing connection, keepalive is a simpler tweak
(although having the option for both is even better...)
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization:
qemu.org |
libvirt.org