Thank you Leo.
This wasnt very well publicised. When changing the upgrade procedure, it should be a bit less buried in the release notes!!
I will give this a try and let you know.
... View more
Hi guys,
We have a whole bunch of Cisco WS-C3650-12X48UQ switches, which were shipped with a now deferred Denali IOS XE release.
Obviously we now need to upgrade. These switches were shipped with their IOS running in INSTALL mode. The upgrade process/commands for this mode is quite straightforward;
- copy image to flash
- run command "software install file flash:file.bin switch #"
- wait, save config, reload
However, the "software" commands are not present in privileged exec, nor configure terminal!
When attempting to run the BUNDLE mode to upgrade, it refuses to boot, and slings me out to the bootloader. I can restart the old IOS using the command "boot flash:packages.conf", but I'm back to square 1.
What is one supposed to do here?
I'd rather not use BUNDLE mode if possible, due to the additional memory constraints - but may have to if no other upgrade method works. Is this yet ANOTHER Denali bug?
Thanks.
... View more
We're using CML 1.2 OVF, which was obtained directly from cisco.com.
Our hypervisor infrastructure has over 16 CPU's and 128GB RAM (it's a scalable blade stack). Plenty of resources there!!
As mentioned in the above post, its ESXi 5.5.0 build 1746018-Custom-Cisco-5.5.1.3
... View more
Hi - I'm having problems attempting to power on the CML VM from my ESXi cluster.
I get an error that says the following: Insufficient resources to satisfy configured failover level for vSphere HA
But the error is bogus. We have enough resources left to power Google on the HA side, and this VM doesnt seem to have any special HA config on it, and only needs a few CPU's and 8GB RAM. The cluster is running about 200 other VM's no problem, and I'm able to build and power on large VM's that have twice the number of CPU's and memory configured.
I've found a list of "fixes" which basically tell me to turn off HA or other silly suggestions. None I'm willing to do to get a single VM powered on.
Is there a bug with this OVF?
We're running ESXI 5.5.0 build 1746018-Custom-Cisco-5.5.1.3
Thanks.
... View more
We're also getting this issue. Running a 3850 with 3 switches in a stack, and seeing a native vlan mismatch on an access port, which is connected directly to a BT 21cn.
It's reporting a mismatch with a switch that is connected on a completely different port, and doesnt actually have a vLAN mismatch at all.
Definitely looks like another bug in the 3850 stack. We are hitting a LOT of these - time to stop buying 3850's I think.
Switch info:
Switch Ports Model SW Version SW Image Mode ------ ----- ----- ---------- ---------- ---- * 1 16 WS-C3850-12XS 03.07.04E cat3k_caa-universalk9 INSTALL 2 16 WS-C3850-12XS 03.07.04E cat3k_caa-universalk9 INSTALL 3 32 WS-C3850-24T 03.07.04E cat3k_caa-universalk9 INSTALL Dec 3 19:44:49.934 GMT01: %CDP-4-NATIVE_VLAN_MISMATCH: Native VLAN mismatch discovered on TenGigabitEthernet1/0/12 (90), with SWITCH05 GigabitEthernet1/0
... View more
Also, just thought I'd point out, we've already configured DHCP option 43 and added a static entry in our DNS servers to direct requests for CISCO-LWAPP-CONTROLLER to the WLC. The problem gets caused by the AP's having invalid network configuration. And it only happend approx 50% of the time - sometimes the AP's come back up with correct network config and there is absolutely no pattern to it that I can see
... View more
Ok, thanks for the tip - will give this a try on the next upgrade I do - have a site with 5 AP's to do this morning so will attempt it with one of those! I still think there is a bug either in the AP firmware or the Upgrade Tool itself. I dont see how else it could get the IP address so drastically wrong. Cheers.
... View more
So are you saying this bug is an actual bug in the Cisco Upgrade Tool? Dont forget, I dont have console to these AP's, so this is being done by telnet and my machine is not a TFTP server (except when running the Upgrade Tool!)
... View more
Hi guys, We're in the proces of upgrading our Wireless AP's from autonomus to LWAP. We're doing the upgrades remotely (we have 100's of sites to do, and it's not possible to be there to console on to each one), using a local computer to upgrade them to LWAP's with the Cisco Upgrade Tool. Having some inconsistent bugs pop up though from time to time... The most common one is that it basically buggers up the IP address after applying the recovery firmware image. It's supposed to keep the IP it had according to the IPFile.txt (for example 172.25.25.4), but it does something very weird... it keeps the last two octets (so 25.4 from the e.g) but replaces the first two octets with a 192.168 address. So the final address in this example after the UpgradeTool has done its work, is 192.168.25.4 - which is compeltely wrong. This causes it to be unable to talk to the WLC (obviously, with bogus IP information it will struggle), and gives us big problems in that it wont ever come up unless someone can console to the AP and fix it. We can usually resolve the issue by munging about on the switch - switching the port it's connected to between access and trunk and doing some shut/noshut cycles seems to do the trick mostly. Is there a way to stop this though? Any help appreciated. Greg
... View more