From: Victor M. Riefkohl (vitinr@bellsouth.net)
Date: Tue Apr 29 2003 - 19:15:13 EDT
Not sure if this has been mentioned also, but have you got with the network
group and ensure that the switches used are compatible with 1GB support. We
have been working on this issue since last year. We have been asking the
network group if the switches are capable to support 1GB, MTU=9000, AUTONeg
=on plus Jumbo frames=on and we have always got a YES for an answer.
Well to make a story short, we just found out last week that the two
switches they solutioned for this project is not capable of doing Jumbo
frames, MTU > 1500 and AutoNeg must be off.
I am dealing with an AIX server with a copper GIGE card and and HP Unix
server with a GIGE fiber card.
Just my 2cents. from our experience so far.
Victor
-------Original Message-------
From: IBM AIX Discussion List
Date: Tuesday, April 29, 2003 5:05:01 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
All switch ports are set to 1000/Full. All ports in Win2K are set to
1000/Full. AIX is set to sense switch port. No errors/dropped packets are
reported with entstat.
-----Original Message-----
From: Chris Leven [mailto:Chris.Leven@GARLOCK.COM]
Sent: Tuesday, April 29, 2003 3:16 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
Just a comment since I haven't worked with the gig cards yet but it sounds
like the old auto negotiate problem we have with the 10/100 cards. Could it
not be coded properly to 1gb full on the AIX/ether switch/ WIN2K?
Christopher Leven
Senior Programmer/Analyst
IBM Certified Specialist AIX Administration
IBM Certified Specialist AIX Support
IBM Certified Specialist RS6000/SP - PSSP 3.1
Garlock Sealing Technologies
An EnPro Industries Company
1666 Division Street
Palmyra, New York 14522
*Voice: 315.597.3034
*Fax: 315.597.7250
*Email: Chris.Leven@Garlock.com
-----Original Message-----
From: Willeat, Todd [mailto:TWilleat@MHP.SMHS.COM]
Sent: Tuesday, April 29, 2003 12:13 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
I added the route and it still didn't work, so I removed it. I rebooted the
Win2K again and I could ping the AIX! Now it gets really weird... I would
get 3 or 4 timeouts and 1 slow ping (700-900ms) out of approx. every 100
ping requests, but it was up and working - for about 6 hours. Then it
stopped again. This time, rather than rebooting, I disabled and reenabled
the network interface in Win2K - and it started working again. Currently, it
is still able to ping. My network guy and I are completely baffled...
-----Original Message-----
From: John Jolet [mailto:john.jolet@FXFN.COM]
Sent: Monday, April 28, 2003 9:09 AM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
for a test, route add -host (man route, i think). to make it stick across
reboots, after your test, do a route delete, then use smit to add it
(communications->tcpip->further configuration->static routes)
-----Original Message-----
From: Willeat, Todd [mailto:TWilleat@MHP.SMHS.COM]
Sent: Friday, April 25, 2003 3:05 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
The traceroute from the Win2K box to the AIX box SHOULD have just one entry.
But I get the same thing as AIX to Win2K - 30 rows of * * * Request timed
out.
No static routes have been added. Where you you put them?
-----Original Message-----
From: John Jolet [mailto:john.jolet@FXFN.COM]
Sent: Friday, April 25, 2003 10:13 AM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
well, when you have multiple interfaces in the box, i think routing does
happen, even if it's local....the traceroute from the win2k box just has one
entry...the aix box, right? I know it sounds silly, but have you tried
adding a static route to the win2k box gated via the correct aix interface?
-----Original Message-----
From: Willeat, Todd [mailto:TWilleat@MHP.SMHS.COM]
Sent: Thursday, April 24, 2003 11:56 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
There shouldn't be any routing since both boxes are on the same subnet...
But, if it helps, here's the traceroute info:
trying to get source for mhp-aquaman
source should be 170.29.112.254
traceroute to mhp-aquaman (170.29.112.30) from 170.29.112.254
(170.29.112.254), 30 hops max
outgoing MTU = 1500
1 * * *
2 * * *
3 * * *
4 * * *
5 * * *
6 * * *
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
-----Original Message-----
From: John Jolet [mailto:john.jolet@FXFN.COM]
Sent: Thursday, April 24, 2003 8:16 PM
To: aix-l@Princeton.EDU
Subject: Re: Strange networking issue
really, really sounds like a routing issue...what does a traceroute from the
aix to win2k box look like?
-----Original Message-----
From: Willeat, Todd [mailto:TWilleat@MHP.SMHS.COM]
Sent: Thursday, April 24, 2003 5:22 PM
To: aix-l@Princeton.EDU
Subject: Strange networking issue
Hi all,
I have a strange network connectivity issue between a p640 box (AIX 4.3.3
ML11) and a Win2K box (SP3). Both boxes are on the same subnet (and same
switch), but are having problems pinging each other. The Win2K box has no
problem connecting to other boxes, and people can telnet to the AIX box. The
Win2K box has 6 GigE connections in an 802.3ad dynamic team. The p640 box
has a SysKonnect dual-port GigE card running in failover mode (as well as
the 2 integrated 10/100s - I have one of these on another subnet and the
Win2K box can ping that just fine...). Looking at the arp tables shows that
each box knows the other's MAC address. My network admin captured some
packets between the 2 and this is what we saw:
Pinging from Win2K to AIX - we saw the packet go from Win2K to AIX, but AIX
did not return any packets.
Pinging from AIX to Win2K - we saw the packet go from AIX to Win2K and saw
the response packet from Win2K to AIX, but AIX did not receive it (request
timed out).
It appears that AIX is dropping the packets from this Win2K box for some
reason (I have several other p640s with the same config that work fine).
However, when I rebooted the Win2K box (after it locked up...), I could then
ping between the boxes. But this only lasted for a day or so, and now it is
not working again.
Has anyone seen anything like this before?
_______________________________________
/ Todd A. Willeat \
| UNIX Administrator |
| Mercy Health Plans |
| 425 South Woods Mill Road |
_ | Chesterfield, Missouri 63017-3492 | _
/ )| (314) 214-2329 / (314) 214-8202 (fax) |( \
/ / | twilleat@mhp.mercy.net | \ \
_( (_ | http://www.mercyhealthplans.com/ | _) )_
(((\ \> \/->___________<-=-=-=-=->___________<-\/ </ /)))
(\\\\ \_/ / \ \_/ ////)
\ / \ /
\ _/ \_ /
/ / \ \
/___/ \___\
This e-mail contains information which (a) may be PROPRIETARY IN NATURE OR
OTHERWISE PROTECTED BY LAW FROM DISCLOSURE, and (b) is intended only for the
use of the addressee (s) named above. If you are not the addressee, or the
person responsible for delivering this to the addressee (s), you are hereby
notified that reading, copying or distributing this e-mail is prohibited. If
you have received this e-mail in error, please contact the sender
immediately.
----------Legal Disclaimer----------
The information contained in this message may be privileged and
confidential, and is intended solely for the use of the named addressee. No
other person is authorized to access, copy or re-use this message (or any
information contained herein). If you are not the intended recipient,
please notify us immediately by replying to this message and delete it from
your computer.
.
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:47 EDT