Hello,
and thanx to all. The only really working solution is to slow down one
of the lines to 10MBit. Not too funny but I have no other possibility.
CU, Dirk.
>> Hello,
>>
>> this is a big cry for help, cause I've no idea what to do.
>>
>> I've an Sun working as a router between two different Fast Ethernet networks.
>> The Sun config is: SunOS 5.6 Generic_105181-11 sun4u sparc SUNW,Ultra-5
>> 320 MB RAM, one additional hme/pci card
>> the interfaces:
>> SUNW,hme0: CheerIO 2.0 (Rev Id = c1) Found
>> SUNW,hme0 is /pci@1f,0/pci@1,1/network@1,1
>> SUNW,hme1: Local Ethernet address = 8:0:20:8f:f8:d3
>> SUNW,hme1 is /pci@1f,0/pci@1/SUNW,hme@1,1
>>
>> Everything works fine. But...
>>
>> if the system gets under heavy load of traffic from one net to the other
>> (ie. round about 10000 packets/sec or more thna 20MBit/sec) one of the
>> interfaces stop to work after a time of roundabout 30-60 seconds.
>> About one or two minutes later the interface is coming back. A typical
>> dump of /var/adm/messages looks like:
>>
>> May 20 08:59:38 wormhole unix: SUNW,hme1: Link Down - cable problem?
>> May 20 08:59:42 wormhole last message repeated 1 time
>> May 20 08:59:43 wormhole unix: SUNW,hme1: Using Internal Transceiver
>> May 20 08:59:43 wormhole unix: SUNW,hme1: 100 Mbps half-duplex Link Up
>> May 20 10:22:43 wormhole unix: SUNW,hme1: Link Down - cable problem?
>> May 20 10:22:46 wormhole last message repeated 2 times
>> May 20 10:22:48 wormhole unix: SUNW,hme1: Link Down - cable problem?
>> May 20 10:22:48 wormhole unix: SUNW,hme1: Using Internal Transceiver
>> May 20 10:22:48 wormhole unix: SUNW,hme1: 100 Mbps half-duplex Link Up
>> May 20 10:23:43 wormhole unix: SUNW,hme1: late collision
>> May 20 10:24:02 wormhole last message repeated 2 times
>> May 20 10:25:48 wormhole unix: SUNW,hme1: Link Down - cable problem?
>> May 20 10:25:52 wormhole last message repeated 2 times
>> May 20 10:25:53 wormhole unix: SUNW,hme1: Using Internal Transceiver
>> May 20 10:25:53 wormhole unix: SUNW,hme1: 100 Mbps half-duplex Link Up
>>
>>
>> The interfaces changes also sometime to the hme0, so I think it's nothing
>> broken on the cards. The late collision is shown sometimes. I've tried both
>> interfaces with full or half duplex, not really changing anything of this
>> behaviour. The seconds before the interface stuck, a netstat -i looks like:
>>
>> netstat -I hme1 10
>> input hme1 output input (Total) output
>> packets errs packets errs colls packets errs packets errs colls
>> 19469 21 9847 35 294 29318 21 29312 35 1296
>>
>> This seems to be normal for me under heavy load. With normal load it looks:
>> 232 0 153 0 0 268 0 191 0 0
>>
>>
>> Any idea, any help or any clue ?????
>>
>>
>> TIA, Dirk.
>>
>>
>>
>> ______________________________________________________________
>> | _|_|_| _|_| _|_|_| _|_|_| _|_|_| _| _| |
>> | _| _| _| _| _| _| _| _| _|_| _|_| |
>> | _| _|_|_|_| _|_|_| _| _| _| _| _| _| |
>> | _| _| _| _| _| _| _| _| _| |
>> | _|_|_| _| _| _| _|_|_| _|_|_| _| _| |
>> |______________________________________________________________|
>> | CAPCom Technologie Beratung | Dirk Boenning |
>> | Entwicklung und Vertrieb GmbH | Tel.: +49 (0)6151/155-900 |
>> | Rundeturmstrasse 6 | Fax.: +49 (0)6151/155-909 |
>> | 64283 Darmstadt / Germany | E-Mail: boenning@capcom.de |
>> |______________________________________________________________|
>>
>>
>
>
______________________________________________________________
| _|_|_| _|_| _|_|_| _|_|_| _|_|_| _| _| |
| _| _| _| _| _| _| _| _| _|_| _|_| |
| _| _|_|_|_| _|_|_| _| _| _| _| _| _| |
| _| _| _| _| _| _| _| _| _| |
| _|_|_| _| _| _| _|_|_| _|_|_| _| _| |
|______________________________________________________________|
| CAPCom Technologie Beratung | Dirk Boenning |
| Entwicklung und Vertrieb GmbH | Tel.: +49 (0)6151/155-900 |
| Rundeturmstrasse 6 | Fax.: +49 (0)6151/155-909 |
| 64283 Darmstadt / Germany | E-Mail: boenning@capcom.de |
|______________________________________________________________|
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:13:20 CDT