[DRBD-user] r0 ok, r1 PingAck did not arrive in time

[Gerald Brandt]

Hi, 

----- Original Message -----

> From: "Chris Tothill" <Chris.Tothill at bristol.ac.uk>
> To: drbd-user at lists.linbit.com
> Sent: Thursday, June 27, 2013 5:42:54 AM
> Subject: Re: [DRBD-user] r0 ok, r1 PingAck did not arrive in time

> Are you using active/active channel bonding? If so I'd suggest trying
> active/passive (or just a single direct connection with no bonding)
> and see if that helps. If in doubt, simplify...

> Incidentally, with gigE my experience is that you don't actually need
> a 'crossover' cable to do direct connection, a normal 'straight'
> cable works just as well. I think this is because gigE uses all 4
> pairs simultaneously in both directions rather than one pair in each
> direction as 100M does.

> Chris

> ---------------------------
> Christopher Tothill
> Zone D Senior IT Support Specialist (Chemistry Unix Support)
> W227, School of Chemistry, University of Bristol
> Tel: (+44 117 33) 17045
> Chris.Tothill at bristol.ac.uk
> ---------------------------

> On 27 June 2013 03:12, Gerald Brandt < gbr at majentis.com > wrote:

> > Hi,
> 

> > I'm running Intel e1000 with crossover cable, and I just noticed in
> > my log files I'm getting this:
> 

> > [3535162.766591] e1000e: eth1 NIC Link is Down
> 
> > [3535168.243278] e1000e: eth1 NIC Link is Up 10 Mbps Full Duplex,
> > Flow Control: Rx/Tx
> 
> > [3535168.243282] e1000e 0000:02:00.0: eth1: 10/100 speed: disabling
> > TSO
> 
> > [3535176.574432] e1000e: eth1 NIC Link is Down
> 
> > [3535178.495165] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex,
> > Flow Control: Rx/Tx
> 
> > [3535214.602022] e1000e: eth1 NIC Link is Down
> 
> > [3535214.602465] e1000e 0000:02:00.0: eth1: Reset adapter
> 
> > [3535239.602540] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex,
> > Flow Control: Rx/Tx
> 

> > Interesting.
> 

> > Gerald
> 

I'm sticking with active/passive, and a double of check of my cable shows it to be a regular one, nothing special about it. 

drbd0 pings/acks just fine, it's only drbd1 that fails, once or twice a day, at irregular times. Will it get worse if I add a third drbd resource to the same box? 

Gerald 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20130627/cb8e6876/attachment.htm>