Re: [Xen-users] Millions of errors/collisions on physical interface in routed Xen-setup??

[Luke S Crawford <lsc@xxxxxxxxx>]

"Rustedt, Florian" <Florian.Rustedt@xxxxxxxxxxx> writes:

> eth0      Link encap:Ethernet  HWaddr 00:XX:XX:XX:XX:XX
>           inet addr:212.XX.XX.XX  Bcast:212.XX.XX.255  Mask:255.255.255.0
>           inet6 addr: xxxx::xxx:xxxx:xxxx:xxxx/64 Scope:Link
>           UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1500  Metric:1
>           RX packets:250388860 errors:8180755 dropped:131050 overruns:0 
> frame:22483
>           TX packets:183684213 errors:0 dropped:0 overruns:0 carrier:0
>           collisions:14726679 txqueuelen:1000
>           RX bytes:231736594230 (215.8 GiB)  TX bytes:43873538477 (40.8 GiB)
>           Interrupt:20 Memory:fd7f0000-fd800000
> 
> Very strange are first the massive collisions, as it is routed and second, 
> that sending has no errors? How can that be possible?


I'd look at netstat -i    my first guess would be that you've got a hardware
problem  (as the tx and rx are done on different pairs, you could have 
bad wires on the rx pair and good wires on the tx pair.)  I see that 
all the time with non-xen hosts.  Cat6 is fairly robust, but it's not 
bulletproof.

And sometimes you get an "A+" data center guy making you network 
cables who thinks it doesn't matter what order he puts the 
wires in so long as it is the same on both ends, and it passes the
test on his $20 at frys "ethernet tester"

uh, also verify you are in full-duplex on the switch and the host; duplex
mismatches will kill you, and while auto-detect is pretty good these days,
and it's been a while since I've seen a duplex mismatch, but especially
if you have old equipment it isn't always perfect.  you shouldn't have 
colisions at all on full-duplex links, even if they are bridged; so something 
is very wrong there.  (you can check duplex with ethtool)   

here is similar output from a host I have (this is xen-bridge, so peth0 
is the physical device rather than eth0.  If I'm not wrong, eth0 is
the physical device when using network-route.)

Now, I get a number of dropped rx packets, but I suspect that is because
this is one of the boxes I set up before I believed the 'dedicate a core
to the dom0' advice.   

[lsc@lion ~]$ uptime
 01:08:44 up 241 days, 23:21,  2 users,  load average: 0.02, 0.01, 0.00

[lsc@lion ~]$ /sbin/ifconfig peth0
peth0     Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF  
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:10386801790 errors:0 dropped:42690346 overruns:0 frame:0
          TX packets:11083748615 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:100 
          RX bytes:1158571105279 (1.0 TiB)  TX bytes:1845366972026 (1.6 TiB)
          Memory:e8180000-e81a0000 

Note, that the box has been up for 2/3rds of a year, and errors and collisions
are both 0;  this is because the physical switch it is on is a switch, and
the connection is full-duplex.  of course, the linux bridge that xen uses
is also a full-duplex connection.



[lsc@lion ~]$ sudo /sbin/ethtool peth0
Settings for peth0:
        Supported ports: [ TP ]
        Supported link modes:   10baseT/Half 10baseT/Full 
                                100baseT/Half 100baseT/Full 
                                1000baseT/Full 
        Supports auto-negotiation: Yes
        Advertised link modes:  10baseT/Half 10baseT/Full 
                                100baseT/Half 100baseT/Full 
                                1000baseT/Full 
        Advertised auto-negotiation: Yes
        Speed: 100Mb/s
        Duplex: Full
        Port: Twisted Pair
        PHYAD: 1
        Transceiver: internal
        Auto-negotiation: on
        Supports Wake-on: pumbag
        Wake-on: g
        Current message level: 0x00000001 (1)
        Link detected: yes



_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users