Cisco Switching/Routing :: Interface Resets Causes Port Channel To Flap Nexus 5010?
Feb 20, 2012
Here is an example of what each switch logs when a server drops offline. Sample logs taken between 5:32am and 5:35am on Feb 20. This particular one was having problems all weekend. Switch #1 encountered over 2000 interface resets. The corresponding VPC port on Switch #2 only had 13 resets.
NEXUS SWITCH #1
2012 Feb 20 05:32:09 q91-sw01-5010 %ETH_PORT_CHANNEL-5-FOP_CHANGED: port-channel10: first operational port changed
from Ethernet1/10 to none
2012 Feb 20 05:32:09 q91-sw01-5010 %ETH_PORT_CHANNEL-5-PORT_DOWN: port-channel10: Ethernet1/10 is down
2012 Feb 20 05:32:09 q91-sw01-5010 %ETH_PORT_CHANNEL-5-PORT_DOWN: port-channel10: port-channel10 is down
2012 Feb 20 05:32:09 q91-sw01-5010 %ETHPORT-5-IF_DOWN_PORT_CHANNEL_MEMBERS_DOWN: Interface port-channel10 is down (
No operational members)
We have a couple of IBM blade chassis with Nexus 4000I 10G blade switch in it. Four 10G ports from each Nexus 4000I switch are bundled into a port channel and connected to Nexus 7K. On the 4k end, one port in the port channel is doing down and coming up as an idividual port. After few minutes it is going down again and coming up as a port-channel member port. It is happening quite offen. [code]
Basically I am trying to use Wireshark to do a packet capture on a Nexus 5010. I want to do a monitor session on on the switch so I can capture from a source port to a destination port on the same switch. I can configure the source port but when I go to configure the destination port I get "ERROR: Eth102/1/4: Configuration not allowed on fex interface". I have tried to reconfigure this port as a switchport but "switchport mode access" command does not take. I don't want to make any changes to any other ports but this one.
We are running nexus 5018 in our DC.What is the difference betwen "channel-group 214 mode active" and " channel-group 216" Any difference?.. because i have problem with this config we are going build a server config?
We have problem with porth channel down.
5K# sh int po71 port-channel71 is down (No operational members) vPC Status: Down, vPC number: 71 [packets forwarded via vPC peer-link]
I am trying to create a port channel between HP servers (4 nic) and two nexus 2k. The server side its a single team with 803.2ad fault taulerence and on the nexus side it have created two port channel (port channel 1 for nexus 2k1 and port channel 2 for nexus 2k2) and made them ACTIVE (channel group mode active)
But when i add a another server on different ports and port channel them the same way as the above server on nexus 2k1 and nexus 2k2, the first server stops pinging. so i have to sht down the first port channel and reopen them - then it works, however it says NO NETWORK ACCESS on the servers (running windows 2008). the only way is to reboot the server i cant be doing this on a production network.
I am having hard time in understanding the hardware port channel resource concept in Nexus 5Ks. Which scenario is considered as a hardware port channel and which is not. According to Cisco documentation, 5548UP switch with layer 3 daughter card only supports 8 hardware port channels, does this mean we can connect only 8 dual homed Fex to those 5Ks.Will a dual homed fex consumes a hardware port channel?
I am a bit confused by the output of 'show run' and 'show run switch-profile' that pertains to a port-channel interface configured in a switch-profile. My main gaol is to find out how can I add/remove the allowed vlans the port-channel (configured as trunk) carries. The setup is like this. I have 2 N5k in vPC domain and Etherner1/11 on both switches is configured as trunk vPC that connects to a core switch. When I issue ‘show run’ for the port-channel and physical interface I get the following output. [code] From above it seems the switch-profile configuration is missing the 'switchport trunk allowed vlan' in the port-channel interface. If want I to remove vlan 30 from the allowed vlan, should I go under the switch-profile mode and remove vlan 30 from the allowed list even though the switch-profile configuration seems to be missing this.
I'm trying to change a vlan on a port-channel but I am getting this error when I apply the change:" error: command is not mutually exclusive",I have done this in the past but I cannot remember exactly what I should type to apply the change.
I'm in the middle of a Nexus 5000 project and recognized today while configuring port-channels, that some of the interface numbers are reserved for internal use.
Is it possible to change or configure which port-channel interface numbers are allocated for internal use by NX-OS?
Unfortunatly I wasn't able to find a solution for this issue in the offical Nexus documentation, the search function of this forum or Google. If I did miss something or didn't look careful enough at the Nexus docs, I'm also happy with RTFM (... fine manual) responses and links to the info :-)
I have a router on which I would like to create a port channel that spans two different Nexus 2248TP switches, and then create a couple subinterfaces on the router's po interface. I've done this successfully on my primary router, but the config doesn't seem to work on the backup router. The primary is running IOS 15, while the backup is running 12.4.25d advanced IP. I do see the trunk come up from the switch side, and my router subinterfaces and po interface show up, however I'm not able to get anywhere from the router. It's not even able to reach other hosts on the same network. As soon as I remove the port channel config and use the physical interface and assign the same address used on the port channel interface, everything works as expected and I have connectivity.
interface Port-channel21 no ip address hold-queue 150 in ! interface Port-channel21.248 encapsulation dot1Q 248 native ip address 100.100.100.1 255.255.255.0
We are planning to implement the following policy map for egress traffic on an Nexus 7000:
policy-map type queuing dd-1p3q1t-8e-out-10G class type queuing 1p3q1t-8e-out-pq1 priority level 1 shape percent 10 class type queuing 1p3q1t-8e-out-q2 bandwidth remaining percent 5 class type queuing 1p3q1t-8e-out-q3 bandwidth remaining percent 5 class type queuing 1p3q1t-8e-out-q-default bandwidth remaining percent 90
We are using two N7K's to which is one N5K connected through a vPC. From the N5k we use a port-channel with 4 * 10G. Two of this four ports are connected to on N7K and the other two are connected to the other N7K. On the n/K's we are using vPC.
My question now are:
1. Where i have to connect the policy map? To the port-channel or on each physical interface?
2. When i have to connect this policy to the port-channel, how does i have to set the shape percent, when i would like to reserve 10% from the 40G? Does i have to set the shape value to 5% on each N7K because vPC?
a Customer ask me to configure a etherchannel between two Switches C3500 XL Version IOS 12.0 follow the first configuration what I done and the output error show me by switch:
I have created 5 new 2 Gig port channels on a 6513 WS-X6516A-GBIC blade connecting to 5 4510R+E switches. 3 of the 5 Port-channels show up/up. 2 show down/down. However, for the 2 showing down, a duplicate Po interface was created with an "A" appended to the name that shows up/up. E.g:
Port-channel26 unassigned YES unset down down Port-channel26A unassigned YES unset up up
Each of the 4510s has a second 2Gig PO to another 6513 with an identical config and all of those come up fine.
how to configure this. I did it in the past but kind of forgot how I did it.I have a stacked 3750 (two physical switches) connecting to a 2960.
I am creating trunk ports with limited access to VLAN 300, 600, and 700.
There is two interfaces connected from the 3750's(one on each physical stack member) to the 2960.I have the physical interfaces configured exactly the same.
Should I keep the configuration on the physical ports and not configure the Port-Channel Interfaces? Do I need to configure port-channel load balancing? Is the channel-group mode sufficient? Goal is to basically create 2 links to the 2960 to double the bandwidth and provide redundancy.
I currenty have a Nexus 5010 connected to a core 3750X switch stack in a VPC trunk using 2 1Gbps links. I want to move this link to 2 10Gbps links without losing connectivity. So I want remove a 1G link and move it to 10G and then once that's up move the other 1G link to 10G hopefully without losing connectivity. So the question is, can I have a 1G and 10G link between the Nexus and 3750s in the same virtual port channel without causing problems?
Our Data Center Switch (5010) rebooted itself today, underneath the captured screen
NX5010-1(config-vlan)# Broadcast message from root (console) (Sun Feb 10 14:22:41 2013):
The system is going down for reboot NOW!
NX5010-1# sh system reset-reason ----- reset reason for Supervisor-module 1 (from Supervisor in slot 1) --- 1) At 740938 usecs after Sun Feb 10 14:22:41 2013 Reason: Reset triggered due to HA policy of Reset Service: nohms hap reset Version: 5.0(2)N1(1)
we have a old Nexus 5010 running version 5.0(3)N1(1)it is in a franckenblock(like frankenstein) ie . we bought the parts and create our own system design before VCE was created. He have since bought VCE V block for our production sites. we use the franc ken block to test before we migrate to the production v block 300 FIX the issue we have is the 5010 will only see two 1 Gb/s SFP fiber modules in the first 12 slots. All these slots are dual 1g/10G. If we add more than two, it claims not to see them. We tried swapping the sfp and using both sfg-GE-MM and GLC-SX-MM moduels, no difference at all 1g sfp are in the first 5 slots. only the first two slots are up.the others say "Link not connected" or "SFP not inserted"all five links are the same SFP and are plugged in other switches.the green link lights are on plugged in SFP, even when the CLI states they are not plugged uni tried both two types of 1 Gb/s SFP. sfg-GE-MM and GLC-SX-MM ,i move the sfp between slot 1,2 and 3,4,5. nothing changed,From "sh int brief" you can see that it can see the sfp ie they are all 1Gfrom "sh int status " it sees 1g but why does it say type 10G?why when i go to the "int e 1/5" and try switch port mode now, it can't mode and trunk at now there (;also how to i stop or clear EU51 %SYSMGR-2-TMP_DIR_FULL: System temporary directory usage is unexpectedly high at 87%.i put as much info in the attached file as i could.
We are looking for some latency in our net and I am trying to check if our STP implementations is running correctly - we have a simple flat network here and no blocked ports here - just two nx 5010 which are interconnected with two uplinks.A remote site - mirrored setup - with 2 10G dark fiber connections - one for each nx5010 - is connection a DR site. I have split the two sites in two stp domains by enabling bpdu-filter on the vpc between the two sites.
I have been running wireshark on the local segment for some time and see the STP RST ROOT announcement every 2. seconds - this is probably normal ? I was looking for some alternate root negotiation packets which would cause the mac tables to be flushed.
I am not able to create more than 256 VLAN in Cisco Nexus 5010 switch. While creating I am getting "No VLAN resources available for VLAN creation" Details below -
Switch model - 5010 Software : NX OS 4.0 (1a)
Error Message: Nexus_5010(config)# vlan 417 ERROR: No VLAN resource available for VLAN creation.
I'm trying to create a vpc between a Nexus 5010 and Nexus 5020 switch. I recently upgraded the software so they are running the same version. I connect get a vpc link. Is there something wrong with my setup? Is a vpc between a 5010 and 5020 even possible? They are connected using a pair of Intel X520's in 802.3AD teaming mode. [code]
My monitoring tool is reporting alerts for high cpu utilization on Nexus 5010.Image is 4.1(3)N1(1) Only command supported on this code is sh proc cpu.The output of which does not really tell what is the current cpu utilization.How do i troubleshoot the cause of high cpu on nexus switches.
if I can use the encrypted port as unencrypted ethernet ports? url.. Each individual port on the Cisco Nexus 5010 switch is numbered, and groups of ports are numbered based on their function. The ports are numbered top to bottom and left to right.There are 20 to 28 ports on the Cisco Nexus 5010 switch, depending on which GEM is installed.
The 20 fixed ports form group 1 and are named 1/port_number. Ports 1 through 16 are unencrypted Ethernet ports. Of these, ports 1 through 8 are 10-Gigabit Ethernet and 1-Gigabit Ethernet-capable ports. Ports 17 through 20 are encryption-capable Ethernet ports.
I am trying to configure get the QLogic 8240 card to work properly in ESXi 5.0. I want to be able to send the iSCSI traffic down the iSCSI portion of the card and use the ethernet portion of the card to do NFS.
Here are my vlans I am working with..
vlan 420 = fcoe vlan 500 = NFS vlan 1000 = iSCSI
I have my interface currenly set as the following on the Nexus 5000.
I found intermittent link down(20~40 seconds average) occurred about 1~10 times every month. SAP reported a lot of active connections are disconnected and I used a batch to ping and found "requested time out" about 30 seconds.And Windows, SQL server, Nexus 5010 do not show any errors. We run cluster and cluster does not fail over.And I don't know which cables or nics cause this issue. When it happened, almost all servers are unreachable. For example, SQL server 1 -> SQL server 2, IBM HS22-1 -> SQL server 1. However, some connections are not dropped sometimes. It varies each time.PS: I run this topology last year without any problems but it started intermittent link down from 2011/1/7. Because there is no errors in Nexus 5010, it is difficult to troubleshoot. Cisco TAC recommended us to implement virtual port channel yesterday. Could I use "errdisable detect cause" to detect what caused the intermittent link down? Is there any error logs or switch parameters/status can use to troubleshoot?
I'm not sure if I'm missing something basic here however i though that I'd ask the question. I recieved a request from a client who is trying to seperate traffic out of a IBM P780 - one set of VIO servers/clients (Prod) is tagged with vlan x going out LAG 1 and another set of VIO server/clients (Test) is tagged with vlan y and z going out LAG 2. The problem is that the management subnet for these devices is on one subnet.
The infrastructure is the host device is trunked via LACP etherchannel to Nexus 2148TP(5010) which than connects to the distribution layer being a Catalyst 6504 VSS. I have tried many things today, however I feel that the correct solution to get this working is to use an Isolated trunk (as the host device does not have private vlan functionality) even though there is no requirement for hosts to be segregated. I have configured:
1. Private vlan mapping on the SVI; 2. Primary vlan and association, and isolated vlan on Distribution (6504 VSS) and Access Layer (5010/2148) 3. All Vlans are trunked between switches 4. Private vlan isolated trunk and host mappings on the port-channel interface to the host (P780).
I haven't had any luck. What I am seeing is as soon as I configure the Primary vlan on the Nexus 5010 (v5.2) (vlan y | private-vlan primary), this vlan (y) does not forward on any trunk on the Nexus 5010 switch, even without any other private vlan configuration. I believe this may be the cause to most of the issues I am having. Has any one else experienced this behaviour. Also, I haven't had a lot of experience with Private Vlans so I might be missing some fundamentals with this configuration.
I have a 5010 that simply won't load any system image. Loads the kickstart image just fine, but once at the Switch(boot)# prompt just give me garbage when I enter "load bootflash:n5000-uk9.5.1.3.N2.1b.bin Restarting system.
Loader Version pr-1.3 loader> dirbootflash: lost+found n5000-uk9-kickstart.5.2.1.N1.1b.bin n5000-uk9.5.2.1.N1.1b.bin n5000-uk9-kickstart.5.1.3.N2.1b.bin n5000-uk9.5.1.3.N2.1b.bin
I attempted to enable jumbo frames on a Nexus 5010 (NX-OS version 4.2(1)N1(1)). I created the policy map below and lost access to the switch.
policy-map type network-qos jumbo class type network-qos class-default mtu 9216
After recovery I see from the logs that all vlans and interface were suspended. I've attempted to look for reasons for a compatibility issue but I am unable to find what is checked and what could have been incompatible. The other troubling thing is the adjacent switch suspended its interfaces too but no change was done there. What I need to look out for so that this does not happen again?
2011 Nov 22 23:43:09 phx-ipcg1dwfcma %ETHPORT-3-IF_ERROR_VLANS_SUSPENDED: VLANs 1,10,601 on Interface port-channel1 are being suspen ded. (Reason: QoSMgr Network QoS configuration incompatible) 2011 Nov 22 23:43:09 phx-ipcg1dwfcma %ETHPORT-5-IF_TRUNK_DOWN: Interface port-channel1, vlan 1,10,601 down 2011 Nov 22 23:43:09 phx-ipcg1dwfcma %ETHPORT-3-IF_ERROR_VLANS_SUSPENDED: VLANs 10 on Interface port-channel508 are being suspended.
I try to connect a cisco blade switch 3020 to Nexus 2232 with etherchannel, but when I connect the second link a obtain flapping on vlans. [code] why the vlans are flapping ? it's something wrong on the config ? [code]
I have configured my Nexus 7018 and 5548 as follow (see diag attached).Both 7000 and 5500 are vPC pair(po1and po2). Now I have created port-chanel between 7018and 5548 as port channel PO3 on 5500. Would it give me 20g bandwidth as PO3?or 10g only uplink to 7018? Do I need to config all four 10g links in PO3 on both 5500 and 7000 switches to achieve max b/w and failover?
We have added GLC-SX-LH module to E1/1 and E1/10 and speed is set to 1000 (1Gig) Now other side of E1/9 and E1/10 are not connected. Port-channel status is down
Po105 1 eth trunk down No operational members 1000(D) lacp
show interface port-channel 105 shows different bandwidth (100) MTU 1500 bytes, BW 100000 Kbit, DLY 10 usec,question is under interface port-channel105 speed 1000 command inserted automatically.Now if we connect otherisde of interface, during port-channel comes up.what will be the port-channel interface bandwidth ? will it change to 2Gig(2000). Why speed 1000 command inserted automatically on port-channel.Since it is migration, we want to be more specific on the port-channel configuration?
