High Ierr Rate on hme0

High Ierr Rate on hme0

Post by Michael D Dav » Wed, 31 Mar 1999 04:00:00



Has anyone seen this before?  It's only happening to about 10 of our ~100
hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis Queue
hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0      0

The hosts vary from Sbus to PCI, server to desktop, so I can't correlate
it to any particular Arch/Daemon/etc.  Hme patches are up-to-date as are
(relatively) kernel patches..

SunOS 5.6 Generic_105181-11 sun4u sparc SUNW,Ultra-1
SunOS 5.6 Generic_105181-11 sun4u sparc SUNW,Ultra-30
SunOS 5.7 Generic sun4m sparc SUNW,SPARCstation-20

All the wiring has been certified to run at Cat5 with 25dB+ of headroom and
most are within 10m of their switches.
The mix of erring machines connect to a mix of Asanti/Cisco/Bay switches
and errors can't be correlated to any particular one..

This is really driving my buggy..

TIA
mike

 
 
 

High Ierr Rate on hme0

Post by Alan L. Stang » Wed, 31 Mar 1999 04:00:00


Hi,

looks like hosts and switches are not running at the same duplex.  As it
sounds like everything is switched, you'll want to make sure everything is
running full duplex.  That probably not the case now.

There are a few ndd command acting on /dev/hme that can interrogate and set
the duplex on the line.  After setting the new duplex, just break the link
(remove and reinsert the RJ45 connector) and the new duplex will be in use.

--


 
 
 

High Ierr Rate on hme0

Post by Lon Stowe » Fri, 02 Apr 1999 04:00:00




Quote:>Has anyone seen this before?  It's only happening to about 10 of our ~100
>hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

>Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis Queue
>hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0      0

>The hosts vary from Sbus to PCI, server to desktop, so I can't correlate
>it to any particular Arch/Daemon/etc.  Hme patches are up-to-date as are
>(relatively) kernel patches..

   Those may not be physical layer errors.  Do you see anything
   odd in netstat -k or netstat -s that might show whether these
   are caused by some IP or driver event?

   If you blast data at the interface and get no packet drop on
   a big file transfer, that would tend to point even more
   towards something other than physical layer.  

   Can't remember if non-supported protocols [e.g. netbios, appletalk,
   netware] also cause these at your patch levels or not.  

 
 
 

High Ierr Rate on hme0

Post by David B. Chorlia » Sat, 03 Apr 1999 04:00:00


With regard to the question below, I recall that there was a
bug in some version of the hme driver that caused it to
to incorrectly classify some events as errors.
This was later corrected.  Anyone experiencing the problem
noted by Mr. Davis should ensure that they have the appropriate
hme software.  (sunsolve.sun.com)




> >Has anyone seen this before?  It's only happening to about 10 of our ~100
> >hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

> >Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis Queue
> >hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0      0

> >The hosts vary from Sbus to PCI, server to desktop, so I can't correlate
> >it to any particular Arch/Daemon/etc.  Hme patches are up-to-date as are
> >(relatively) kernel patches..

>    Those may not be physical layer errors.  Do you see anything
>    odd in netstat -k or netstat -s that might show whether these
>    are caused by some IP or driver event?

>    If you blast data at the interface and get no packet drop on
>    a big file transfer, that would tend to point even more
>    towards something other than physical layer.

>    Can't remember if non-supported protocols [e.g. netbios, appletalk,
>    netware] also cause these at your patch levels or not.

--
David B. Chorlian
Senior Scientific Programmer, Neurodynamics Lab, SUNY/HSCB
voice: 718-270-2231; fax: 718-270-4081

 
 
 

High Ierr Rate on hme0

Post by Michael Dav » Fri, 09 Apr 1999 04:00:00


Thanks..  netstat -k shows most all of the errors on hme0 to be CRC
errors.  I've also determined that all the affected hosts are directly
connected to (or thru a dumb switch) our CISCO 2926s.  Again, no permutation
of Autonegotiate or hard-coding between the host/switch makes a difference
and serveral other machines also directly connected to the CISCOs don't
show problems..  We recently moved one of the affected hosts onto a
Torrent Router that hangs off of the CISCO 2926, and the affected host
stopped showing errors!

Should I take up this problem with CISCO?

thanks
mike





>>Has anyone seen this before?  It's only happening to about 10 of our ~100
>>hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

>>Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis Queue
>>hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0      0

>>The hosts vary from Sbus to PCI, server to desktop, so I can't correlate
>>it to any particular Arch/Daemon/etc.  Hme patches are up-to-date as are
>>(relatively) kernel patches..

>   Those may not be physical layer errors.  Do you see anything
>   odd in netstat -k or netstat -s that might show whether these
>   are caused by some IP or driver event?

>   If you blast data at the interface and get no packet drop on
>   a big file transfer, that would tend to point even more
>   towards something other than physical layer.  

>   Can't remember if non-supported protocols [e.g. netbios, appletalk,
>   netware] also cause these at your patch levels or not.  

 
 
 

High Ierr Rate on hme0

Post by Alan L. Stang » Fri, 09 Apr 1999 04:00:00



> Thanks..  netstat -k shows most all of the errors on hme0 to be CRC
> errors.  I've also determined that all the affected hosts are directly
> connected to (or thru a dumb switch) our CISCO 2926s.  Again, no permutation
> of Autonegotiate or hard-coding between the host/switch makes a difference
> and serveral other machines also directly connected to the CISCOs don't
> show problems..  We recently moved one of the affected hosts onto a
> Torrent Router that hangs off of the CISCO 2926, and the affected host
> stopped showing errors!

> Should I take up this problem with CISCO?

I would do two things at once.  Take it up with Cisco (which is probably going
to be a frustrating experience) and also check out the other systems on the
switch.  Are all the systems attached to the switch Suns?  If there is a PC, try
removing it just in case it's misbehaving somehow.

What if you swap switch ports between an errorprone and error free system?  In
other words, do the errors occur on certain switch ports only?

--

 
 
 

High Ierr Rate on hme0

Post by Paul Borgerman » Sat, 10 Apr 1999 04:00:00


We have the same "problem" here on our Solaris 2.6 boxes,
but surprisingly the Ierr rate drops to zero if the interface is set to
promiscious mode with snoop. An identical machine (hardware) with
Solaris 2.5.1 on the same swith does not show this high Ierr rate.

Can you reproduce this ?

Cheers



> > Thanks..  netstat -k shows most all of the errors on hme0 to be CRC
> > errors.  I've also determined that all the affected hosts are directly
> > connected to (or thru a dumb switch) our CISCO 2926s.  Again, no permutation
> > of Autonegotiate or hard-coding between the host/switch makes a difference
> > and serveral other machines also directly connected to the CISCOs don't
> > show problems..  We recently moved one of the affected hosts onto a
> > Torrent Router that hangs off of the CISCO 2926, and the affected host
> > stopped showing errors!

> > Should I take up this problem with CISCO?

> I would do two things at once.  Take it up with Cisco (which is probably going
> to be a frustrating experience) and also check out the other systems on the
> switch.  Are all the systems attached to the switch Suns?  If there is a PC, try
> removing it just in case it's misbehaving somehow.

> What if you swap switch ports between an errorprone and error free system?  In
> other words, do the errors occur on certain switch ports only?

> --


--
----------------------------------------------------
Paul Borgermans            tel: +32-14-333227
Dept W&D                   fax: +32-14-323553

Boeretang 200              www: http://www.sckcen.be
B-2400 Mol
Belgium
----------------------------------------------------

  pborgerm.vcf
< 1K Download
 
 
 

High Ierr Rate on hme0

Post by Bryan Bahnmille » Sun, 11 Apr 1999 04:00:00


> Thanks..  netstat -k shows most all of the errors on hme0 to be CRC
> errors.  I've also determined that all the affected hosts are directly
> connected to (or thru a dumb switch) our CISCO 2926s.  Again, no permutation
> of Autonegotiate or hard-coding between the host/switch makes a difference
> and serveral other machines also directly connected to the CISCOs don't
> show problems..  We recently moved one of the affected hosts onto a
> Torrent Router that hangs off of the CISCO 2926, and the affected host
> stopped showing errors!

> Should I take up this problem with CISCO?

> thanks
> mike


> >>Has anyone seen this before?  It's only happening to about 10 of our ~100
> >>hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

> >>Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis Queue
> >>hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0      0

> >>The hosts vary from Sbus to PCI, server to desktop, so I can't correlate
> >>it to any particular Arch/Daemon/etc.  Hme patches are up-to-date as are
> >>(relatively) kernel patches..

  We've seen this with our E3000's. We have one E3000 running with high
Ierrs on its hme interfaces, the other running with no errors. They are
both connected to the same Cisco 5500. They are both running 2.6, but
with different patch levels. I'm at home now and can't check, but the
one running with no errors has a patch installed whose documentation
refers to excessive input errors - check Sunsolve to be sure.

  Bryan

 
 
 

High Ierr Rate on hme0

Post by Ashton Trey Bele » Thu, 15 Apr 1999 04:00:00


...
hosts that are running hme0 at 100Mbs/Full-duplex.  A sample:

Name  Mtu  Net/Dest   Address        Ipkts  Ierrs Opkts  Oerrs Collis
Queue
hme0  1500 x.x.x.0    x.x.x.3     44756446 309740 53221694 0     0    
0

Try something like this at boot:
ndd -set /dev/tcp tcp_old_urp_interpretation 1
ndd -set /dev/hme instance 0
ndd -set /dev/hme adv_100T4_cap 0
ndd -set /dev/hme adv_100fdx_cap 1
ndd -set /dev/hme adv_100hdx_cap 0
ndd -set /dev/hme adv_10fdx_cap 0
ndd -set /dev/hme adv_autoneg_cap 0
        That will keep the hme interface from trying to autonegotiate 100 full
duplex which it is not smart enough to do properly.  As written above,
it will hard-wire hme0 to 100 full duplex, but by moving the 1's and 0's
(dont you feel we do that all day, though :) you can change to 100 half
etc etc...
        For whatever reason, once it no longer attempts the autonegotiate, it
will more than likely work properly.

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ashton Trey Belew
Wesleyan University
Phone: 860 685 3142

http://veggie.wesleyan.edu/trey/
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
We are all in the gutter, but some of us are looking at the stars.
                -- Oscar Wilde

 
 
 

High Ierr Rate on hme0

Post by Michael Dav » Sat, 17 Apr 1999 04:00:00


promiscious mode does not change the error rate on my machines..

thanks
mike



>-=-=-=-=-=-

>We have the same "problem" here on our Solaris 2.6 boxes,
>but surprisingly the Ierr rate drops to zero if the interface is set to
>promiscious mode with snoop. An identical machine (hardware) with
>Solaris 2.5.1 on the same swith does not show this high Ierr rate.

>Can you reproduce this ?

>Cheers



>> > Thanks..  netstat -k shows most all of the errors on hme0 to be CRC
>> > errors.  I've also determined that all the affected hosts are directly
>> > connected to (or thru a dumb switch) our CISCO 2926s.  Again, no permutation
>> > of Autonegotiate or hard-coding between the host/switch makes a difference
>> > and serveral other machines also directly connected to the CISCOs don't
>> > show problems..  We recently moved one of the affected hosts onto a
>> > Torrent Router that hangs off of the CISCO 2926, and the affected host
>> > stopped showing errors!

>> > Should I take up this problem with CISCO?

>> I would do two things at once.  Take it up with Cisco (which is probably going
>> to be a frustrating experience) and also check out the other systems on the
>> switch.  Are all the systems attached to the switch Suns?  If there is
>a PC, try
>> removing it just in case it's misbehaving somehow.

>> What if you swap switch ports between an errorprone and error free system?  In
>> other words, do the errors occur on certain switch ports only?

>> --

>--
>----------------------------------------------------
>Paul Borgermans            tel: +32-14-333227
>Dept W&D                   fax: +32-14-323553

>Boeretang 200              www: http://www.sckcen.be
>B-2400 Mol
>Belgium
>----------------------------------------------------

>-=-=-=-=-=-
>[Attachment type=text/x-vcard, name=pborgerm.vcf]
>-=-=-=-=-=-

 
 
 

1. High Scan Rates, High IO Wait Question

Hello,

I'm an Oracle DBA who is relatively new to AIX, and I have a server
tuning question regarding proper settings for vmtune.

Server: H60, 4 CPU, 3 Gig RAM, 2.4 GIG paging (3x800M), AIX 4.3.3
Application: Oracle 8.1.6 servicing PeopleSoft Financials 7.5

The machine is a dedicated Oracle server.

Symptoms:
Currently, vmtune -p 15 -P 60.  Paging, as reported by vmstat, has
almost disappeared.  However, it appears that when Oracle dedicated
server processes are started, the scan rate goes through the roof, I see
spikes at around 14000 to 16000 runing vmstat 120.  We also see very
high IO wait time that I believe is related to VMM activity.

I think the other vmtune parameters are default settings, -f 121 -F 128.

What can I do to reduce the IO wait and the high scan rates, or, are
scan rates this high normal for this environment?

I've looked long and hard at various manuals and web pages, but I can't
seem to find specific settings to fix this problem.

Thanks,
Nick

2. xemacs and tset?

3. How can I sustain a high transfer rate ?

4. Time Synch

5. PPP lost packet rate very high

6. Problems installing COM3 serial port

7. Using Linux to Capture Packets on High Speed/Rate Network

8. RedHat vpop3d Problem

9. High collision rate - what to do

10. High Sun Ray 1 failure rate

11. SB PnP working, but DMA time out when trying to play HIgh sample , rate file.

12. Qs on attack by high rate process creation

13. using Bluetooth for high-data-rate comms