I try to get an EtherChannel to a Sun with four GigaSwift Adapters and Sun Trunking 1.3/Solaris 9 up and running. The Sun guys complain that they have freezes in their NetBackup tool. I don't see any log messages about the interfaces going down nor do I see any errors on the interface.
Does anyone have a running config with Sun and 6500 or are there any known problems?
We have Etherchannel configured to SUN 25K running Solaris 9/Trunking 1.3 patch/upgrade. We are trunking between 6509 with 6748-GE-TX modules and GigSwift interfaces using LACP and IP source/dest (XOR algorithm) load-balancing. The channel groups are configured as trunks. The switch IOS is 12.2.18. We have not experienced any performance issues.
Q's - What is the line card(s)? What is the OS version of switch? Are you using LACP, or are the etherchannels forced? What type of load balancing are you doing? Can you provide "show etherchannel summary" or show "LACP NEIGHBOR DETAIL"?
thanks a lot for your response. I wrote the post in a hurry so almost anything was wrong that I wrote. The switches I have (shame on me) are 4500s with WS-X4424-GB-RJ45 with 12.1(20)EW2. I use LACP in active mode and in my opinion it works fine as far as missing error messages are concerned.
Due to the fact that we had to revert to a running environment the ether channels are not configured at the time so i cannot provide you the show commands that you requested.
I do not know where the Sun administrator has the lacp keyword from but it more or less looks like 'active' to me. How did you configure your channel? active mode as well or anything different? Does the nettr make sense to you?
One thing about the Sun Trunking PDF: on page 34 it states that Sun Trunking 1.3 does not support the LACP and Marker protocols. This puzzles me because in patch 115853-06 on sunsolve they talk about dynamic LACP big time.
I hope that the Sun admin will have time tomorrow to do some more testing. Then I'll let you know about the results.
Yeah, the document (PDF) was pre LACP. The patch would make LACP configuration possible.
The NETTR looks fine. However, things to note are the policy and LACP settings. Policy is to determine load-balancing. It looks like you are using the default MAC option. And, yes LACP is set to ACTIVE (we are using the same on the server side, but the switch ports are set to passive).
For load-balancing refer to page 31 of PDF. There are 4 options to consider. Each with pros-and-cons depending on your environment. We prefer IP-src-dest (policy=4).
An important thing to note is how your Sun boxes are configured. Some SUN servers, by default, use a global MAC address for all interfaces. This could result in the symptoms you described. This must be modified for trunking to work. Check your servers local-mac-address setting. It should be set to TRUE.
Without knowing how your switches are configured I would start with these server settings; local-mac-address=true, and on NETTR make policy=4 (just because the MAC hash has been known to create problems that I don't know if they have been addressed in the latest patch. It has to do with packets being received out of sequence ,or something along those lines).
Thanks to the info in this thread we've got sun trunking v1.3 working (on a 280r with sol9 & gigaswift 4 port card) hooked up to a 4500 switch.
One slightly puzzling issue is that I setup l2 port-channels on the switch without any specific cisco trunking just access ports, whereas sun always refers to trunking. I guess this is just different terminology?
the trunk term is a multivalue term in networking. Sun (and other vendors) use it to specify the bonding of multiple physical interface into one logical interface. Cisco uses the term EtherChannel for the same method.
When Cisco talks about trunking the refer to the mechanism that others descrive with tagging. It means that you run multiple VLAN over the same interface. It does not depend whether its is one cable or if it is an EtherChannel connection. You will recognize trunking on a Cisco switch if you find config statements like 'switchport mode trunk' and/or 2 switchport trunks encapsulation dot1q|isl'.
To make the picture complete there's another usage of the t-word in voice: if voice engineers talk about trunks the refer to as one connection between two voice switches. This can be one analog line (FXS, FXO, E&M) or one digital timeslot on an E1/T1 link. It is often used in capacity measurement situations where you have to find out how many trunks you need in order to get the most out of your telephone system with the least possibility of an overloaded system (at the lowest possible price, to be correct).
BTW: what line cards are you having in the 4500? are you using 4 links to the Sun's? Are you by accident using NetBackup on the Sun's?
This is actually a pretty cool feature, i didn't even know it existed until I was looking for a solution to advertise a subnet (prefix in BGP talk), only if a certain condition existed. This is exactly what conditional advertisements does
j ai une question j ai achete un routeur cisco 887VA-k9 , je le configuré avec la configuration ci- dessous
si je le lier avec mon pc portable sur l un de ses ports directement ça marche toute est bien ( la connexion internet + m...
Attached policy provides CLI access to the Cisco 4G router over text messaging. Two files are in the attached .tar file:
2. PDF with instructions on how to load and use the .tcl file.