[libvirt-users] livemigrate problems

saint at eng.it saint at eng.it
Wed Jan 5 13:20:31 UTC 2011

>>>>> "CC" == saint  <saint at eng.it> writes:

>>>>> "CC" == saint  <saint at eng.it> writes:
CC> I can livemigrate w/o problems between any two from red9, red10
CC> and red11, I can livemigrate from either red9, red10 or red11 to
CC> red3 but I can't livemigrate to any other blade from red3.

I TCP-dumped the  dialog from both a succesful and  a failed migration
(as I said,  I have a couple of problematic  hosts and other perfectly
working ones).

When the  migration succeeds, I have  data transmitted to TLS  port on
the target  machine and  to one  of the  ports that  must be  open for
migration with this pattern:

first data are sent  to the TLS port, then to the  other port (and the
communication ends  with a reset), then  other data ended by  a FIN to
the TLS port on the target machine. The reply to the FIN is a RESET.

I checked, it is not a problem  with the certificates: I had a machine
raising  such  problems  because  had  the clock  set  to  before  the
beginning of the vvalidity of the client certificate.

I checked libvirtd configuration, and got nothing strange.

The machine that fails to migrate can contact the intended target:

virsh -c qemu://remotehost/system list --all 

works perfectly.

I think I am going to fill a bug report...

ing. Gian Uberto Lauri
Ricercatore / Reasearcher
Divisione Ricerca ed Innovazione / Research & Innovation Division
GianUberto.Lauri at eng.it

Engineering Ingegneria Informatica spa
Corso Stati Uniti 23/C, 35127 Padova (PD) 

Tel. +39-049.8283.538         | main(){printf(&unix["\021%six\012\0"], 
Fax  +39-049.8283.569		  |    (unix)["have"]+"fun"-0x60);}   
Skype: gian.uberto.lauri	  |          David Korn, AT&T Bell Labs             
http://www.eng.it 			  |          ioccc best One Liner, 1987             

More information about the libvirt-users mailing list