Cisco Switching/Routing :: 3750x Stack - LACP Waiting To Be Aggregated Takes Too Long
Oct 26, 2012
I have configured a portchannel with 3 ports - lacp active mode I want to pxe boot esx server with Hp Proliant DL 360 G7.
During Pxe Boot the physical Interfaces must be in (I) state - standalone.
The server receives via dhcp the ip but after getting the ip the link goes down for a short time and the switchport goes down and then in (w) state - waiting to be aggregated. Now the server tries to load via tftp the image but the server gets a timeout because the switchport is in state (w). It takes a long time til the switchport comes back in standalone mode (I).
Is it possible to modify the time how long the switchport is in the waiting to be aggregated state on 3750x ?
Would the right command be lacp rate-fast? But this command is not available on 3750x, only 6000 series.
We will be converting two 6500s to VSS. Each chassis has a Sup2T module, fwsm, 6908-10G blade, wism1 blades and two or three 6748-SFP blades. I was wondering how long the conversion process takes? If I remember correctly at the 2012 networkers lab it took maybe 10 minutes while using Sup720s. The instructor mentioned that with Sup2Ts it would convert a lot faster.
I have 3 switches in my 3750 X stack. I am getting the following error in the logs: %PLATFORM_STACKPOWER-4-PRIO_CONFLICT: Switch 1's power stack has conflicting power priorities. Not sure what I should do next. Below is my Stack-power configuration:
Power stack name: Powerstack-1 Stack mode: Power sharing Stack topology: Ring Switch 1: Power budget: 735 Low port priority value: 20 High port priority value: 10 Switch priority value: 2 Port 1 status: Connected Port 2 status: Connected Neighbor on port 1: c464.1308.7900 Neighbor on port 2: e8b7.4827.e680 Switch 3: Power budget: 729 Low port priority value: 22 High port priority value: 13 Switch priority value: 4 Port 1 status: Connected Port 2 status: Connected Neighbor on port 1: 2894.0fd9.9f00 Neighbor on port 2: 2894.0fd8.e200 Switch 2: Power budget: 690 Low port priority value: 21 High port priority value: 12 Switch priority value: 3 Port 1 status: Connected Port 2 status: Connected Neighbor on port 1: 2894.0fd8.e200 Neighbor on port 2: c464.1308.7900
Each switch has its own 715W power supply in it and all check out OK.
I am having an issue on a Cisco 3750 stack where when the stack master is rebooted, all my lacp port-channels drop and then come back up again. After doing some investigation It seems that it is happening because of lacp using the stack master mac-address as part of the system-id, so when the stack master reboots, the stack mac changes. I see that there is the command: stack-mac persistent timer 0
There is this warning about using this command:
When you configure this feature, a warning message displays the consequences of your configuration. You should use this feature cautiously. Using the old master MAC address elsewhere in the domain could result in lost traffic.
My question are:
Are there any other consequences to using this command (apart from moving the switch/mac to another location in the network)It mentions 'If the entire switch stack reloads, it acquires the MAC address of the master as the stack MAC address' Is this still the case if you have the stack-mac persistent timer to 0? Does using channel-group mode on for the port-channels still use the same mechanism of having a system-id? (Will the channels flap using 'mode on' when rebooting the stack master.
We have two dedicated racks for our DMZ. We need to connect them to a firewall. We do all the cabling inside the rack, so each of them must have a top of rack.Problem is, switch in rack 2 will be connected to rack 1 and then to firewall, so if switch 1 fail, we lose both of them.
I would love to stack them like 3750 but I don't think there is stack-wise cable long enough to connect two racks (and I suppose this length isn't supported by technology).We have only one port left in the FW.
I'm trying to review a QoS setup, and I'd like to make sure I fully understand the current setup before I change anything. I'm seeing output drops on two different queue-thresholds, but not sure how packets are making it to one of the queues.
Switch Version CORE#show ver Cisco IOS Software, C3750E Software (C3750E-UNIVERSALK9-M), Version 15.0(2)SE, RELEASE SOFTWARE (fc1) System image file is "flash:/c3750e-universalk9-mz.150-2.SE/c3750e-universalk9-mz.150-2.SE.bin" cisco WS-C3750X-24 (PowerPC405) processor (revision A0) with 262144K bytes of memory. Switch Ports Model SW Version(code)
I can find queue4-threshold3 in the mappings, but how are packets getting mapped to queue2-threshold1? The priority queue is disabled for this interface, so I'm not sure how this queue is dropping packets, according to the maps nothing is mapped to 02-01.
This 3750X stack (3 members) has some 100% cpu spikes. Is this bad enough to start tracking down the cause(s) or is this somewhat typical and not a reason for concern? I just took this reading. How can you detemine when the 100% and 80% spikes occurred?
I recently created a stack using (2) 3750x switches. I have three vlans on the stack (1,105,241) Vlan 105 is configured on 6509 core switches with multicasting and are connected to the stack via gigabit fiber. It is working well with clients on the Master or SW1, however clients on SW2 do not participate in multicasting Any client pc that is connected to SW2 vlan 105 does not show up in the "sh ip igmp snooping group" command. I can statically assign a client on SW2 to the mcast groups but but they fail to register on their own. I can take the cable connection from SW2 and plug it into an SW1 Vlan 105 port and it immediately becomes a member of the groups. I can then connect it back to the SW2 port and it disappears from the group membership. IGMP and PIM are configured with the defaults.
Seeing very high CPU on one of our 3750X stacks. Usally its around 30% but recently its 70, 80, 90 or maxed at 100% during the main day.IOS version c3750e-universalk9-mz.122-55.SE1.bin.Has a ten gig uplink connected to a nexus 7k and another ten gig connected to another 7k. Running Layer 3 and EIGRP on the 3750X.Is this normal for the CPU?
#sh processes cpu history 5555555554444455555555555555588888666669999888885555577777 8888000001111111111222227777755555666662222222229999977777 100 90 ***** **** 80 ***** ********* *****
I have two WS-C3750X-48PF-S in stack, but every one has just one power supply. In the package of the switches is included one stack power supply cable.does it make sense to use stack power cable if every switch has just one power supply ??
Switch 2 has an interface configured in RSU Standby.Switch 3 has an interface configured in RSU active. The upgrading sequence is : Switch 2, Switch 1, Switch3.The aim of the test is to prove that there is no loss of IP connectivity throught the RSU process.During the RSU process, there are ping between all the computers.
IOS starting version : 15.0-1.SE3 IOS upgrading version (with OVERWRITE) : 15.0-1.SE3 (same version)
1rst step : Switch 2 reboot and come back master but isolated from original stack (switch 1 and 3)
2nd step : Switch 1 reboot, and come back with switch 2 into the upgrading stack.
Issue : at this step :
- Switch 1 put its interface (IP 192.168.5.254) up :
- Switch 3 has not reboot and still has its interface (IP 192.168.5.254) up
=> we have some OSPF adjencies and mac flapping issues : switch 1 and 3 are trying to mount there OSPF adjencies on the same IP. We loss our IP connectivity since step 3 is not over.
3rd step : Switch 3 reboot and come back up in the upgrading stack. The IP connectivity is back, and the RSU process is done.Is that a normal behaviour ?How can we avoid to loose IP connectivity during RSU process ?
I've a situation where I need to add a new 3750x to a existing stack of 3750. [code] When I tried to stack them together, I get a version mismatch error. Is this because of the difference in SW Image? What are my options next ? My ultimate goal is to make the new switch stack correctly with the exisitng switches.
IOS upgrade tried on the stack of 2XWS-C3750X-24T-S , Upgrade was tried to support WCCP on the platform. Current version is Cisco IOS Software, C3750E Software (C3750E-UNIVERSALK9-M), Version 12.2(58)SE1, Not able find any IP Plus ios in the site ? Only available are IP base and Universal K9 , Tried upgrading to c3750e-universalk9-tar.122-58.SE2.tar by using stack upgrade method as per document [URL]
Upgrade fails with error :
%Error opening flash:update/info (No such file or directory) ERROR: Image is not a valid IOS image archive.
Why current ios is not supporting WCCP , I hope it is there in the featureset. ( tried "sdm prefer routing" but no success) Does c3750e-universalk9-tar.122-58.SE2.tar will give me wccp support ? What is the issue with my upgrade and how do I do a successfull upgrade ?
We recently configured a stack of four 48 port 3750-x switches . We are noticing high CPU usage. "Hulc LED process" seems pretty high. This has coincided with VMware servers getting slow and non-responsive at times, perhaps a coincidence, not sure. Below I provided some outputs about how to diagnose it.
if I read the Datasheet of Catalyst 3750X-Series-Switches it is possible to connect a new X-Switch to an existing and old Catalyst 3750-Series Stack.What kind of requirements are needed? Only same IOS-Version in the hole Stack and if possible same Feature-Set? .... like in a normal NOT mixed Stack?
Our customer has a stack of 6 3750G switches. They want to add another switch to the stack. I tried to quote another G, but I got a warning that the G series is EOS January 30,2013. Can I add a 3750X to the stack? I know that no matter what switch I get, I will have to upgrade the IOS in the stack, as the stack is running 12.2(25)SEE2.
I have a 3750X 24 port switch (with NM-1G network module) running IOS 15.0(1). Is there any benefit or reason to plug in the included Stack Wise cable and loop it back to itself in a single switch installation?
I don't see any recommendation in the documentation. The data sheet indicates a single switch is a non-blocking device so I'd think there's no bandwidth advantage like there is when connecting an actual multi-switch stack and needing to close the loop for the full 32 Gbps stack bandwidth.
Is that possible to configure a stack between then?
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Tenho um cisco WS-C3750G-24TS com o IOS flash:/c3750-ipservicesk9-mz.1 22-25.SEB4.bin" Adquiri um novo cisco WS-C3750X-24 com o IOS flash:/c3750e-universalk9-mz.150-1.SE/c3750e-universalk9-mz.150-1.SE.bin.
I have two 3750X in a stack. Each 3750X has one 10Gb uplink to primary core1 and one 1Gb uplink to backup core2 configured in 2 etherchannel groups. The member switch has bad ports, so I need to replace it. I have spare 3750X unit. Is it possible to connect the spare unit to the stack and then move the RJ-45 cables from the member with bad ports to the new member? That way I can minimize servers downtime to just a few seconds..If that would work, then the last step would be to move the 10Gb networking module from old member to the new one, as well as the secondary PSU. Is it doable? Also according to the documentation, all I need to add new member is matching IOS version, correct?
I have two 3750G switches stacked together however it's not in a full ring configuration:
Stack Ring Speed : 16G Stack Ring Configuration: Half Stack Ring Protocol : StackWise
And as shown below it's obvious which cable is not working:
show switch neighbors Switch # Port 1 Port 2 -------- ------ ------ 1 2 None 2 None 1
It's not flapping - it's like the cable isn't there at all. I'm going to try reseating the cable in there and perhaps an all out replacement if necessary. My question - is there any threat of a reboot of the switch stack or of partitioning of the switch stack by trying to get the 2nd cable working?
i am seeing very high utilization on a random basis on the stack. and the logs indicate me that there is a pattern where there is a stack power cable shows inserted (which was never unplugged) followed by sudden spike in the utilization.
001018: *Mar 6 16:21:22.138: %PLATFORM_STACKPOWER-6-CABLE_EVENT: Switch 4 stack power cable 1 inserted 001019: *Mar 6 18:18:37.982: %SYS-1-CPURISINGTHRESHOLD: Threshold: Total CPU
I have a 3750X stack of 6 switches, I have a terminal server console server plugged into the console ports of all 6 switches. I am having a strange issue where I can connect to the master via the console, get a login prompt then I can log in, but when I try and connect to any of the members then I get a prompt "Press RETURN to get started" When I push return, it just loops back to the same prompt again:
switch-6 con0 is now available Press RETURN to get started.
switch-6 con0 is now available Press RETURN to get started.
I can connect to the stack via SSH fine etc. I am running 15.0(1)SE2. The stack is configured for TACACS+ auth via Cisco ACS.
Switch# Role Mac Address Priority Version State ---------------------------------------------------------- *1 Master f0f7.55ba.2700 15 1 Ready 2 Member f0f7.5541.da00 14 1 Ready 3 Member a493.4c1e.4b80 13 1 Ready 4 Member f0f7.5585.cb80 12 1 Ready 5 Member a44c.1120.8e80 11 1 Ready 6 Member a493.4cc0.fa80 10 1 Ready