Template deployment bug in Prime Infrastructure 3.1.0

[powys]

There appear to be a few odd bugs in CPI 3.1.0 that weren't there in 3.0.3. Here's what I was getting when trying to deploy an updated preauthentication ACL to four WLCs (one 5508 on 7.0.252.0 plus two 5508s and an 8510 on 8.0.121.0). The update was a simple one; two new rules to allow inbound and outbound traffic for an external VPN IP address.

Firstly, the 5508 running 7.0.252.0.

With Administration -> System Settings -> Inventory -> Configuration -> Backup Device Configuration ON, deployment would fail with the following error:-

Archive before deployed is failed. Check error details from Job Dashboard by manually scheduling archive for the device. To continue deploy in these scenarios, disable Configuration Archive Setting from Administration --> System Settings --> Configuration

The configuration on the device itself didn't change.

With it OFF, the error changed to the following:-

SNMP operation to Device failed: COMMON-1: Some unexpected internal error has occurred. If the problem persists please report to the Tech Support.

Checking the controller through its own web interface revealed an Empty ACL with the wrong name (in our case PCC instead of PCC Preauthentication ACL).


The devices running 8.0.121.0 had a different error. Whether the above setting was ON or OFF:-

SNMP operation to Device failed: Get failed after set

On the device, Any-Any "Deny" rules appear in place of the new rules that were added to the template, plus a further Deny rule at the end. If these are left in place they block the guest wifi landing page, so they needed to be removed from the controllers using their own web interfaces.

When attempting a Configuration Archive from the device running 7.0.252.0:-

Error occured while communicating with the device. Check device credentials and SSH/telnet reachability.

Meanwhile, on 8.0.121.0 the operation was successful.

All SNMP and SSH credentials on controllers of both versions verify OK when tested and I can SSH into the devices using Tera Term.

I've now switched back to the previous CPI 3.0.3 VM and the above operations work properly. The 3.1.0 VM was a fresh build with the config backed up on 3.0.3 and restored to 3.1.0.

[RQ-Support-IT]

When attempting a Configuration Archive from PI 3.1.0 for WLC 5508 running version 8.0.121.0  we are having this error message too.  Please advice

 "" Error occured while communicating with the device. Check device credentials and SSH/telnet reachability.""

[Marco Matera]

I have the same problem with AP Groups. I have 3 WLC Controllers with the following versions and i get the following error when i deploy a AP Group template which already exists on WLC with several SSID's.

WLC 8510 with Version 7.4.121.0       --> Deploy successfull 

WLC 8510 with Version 8.0.120.0       --> SNMP operation to Device failed: Get failed after set

WLC 5508 with Version 8.0.120.0       --> Deploy successfull

After this failure (SNMP operation to Device failed: Get failed after set) there is no SSID in the AP Group on the controller and i have to push the AP Group again and then the deploy is successfull and all SSID's are correct shown on the WLC in the AP Group.

We have this Problem since Prime 3.0 now we use the newest Version 3.1.3 and the problem still exists.

[powys]

A possible fix for "Archive before deployed is failed." errors is to increase the telnet/SSH timeout.

I've just returned to our 3.1.x build (upgraded to 3.1.6 this time) and got errors on our 5508 controllers but not the 8510. Attempting a scheduled archive would give a "Function run timed out 60000ms" error.

Performing a show run-config on the controller gave a clue; the config would freeze for a short while partway through running (at the start of the 802.11b config in our case), Prime may have encountered the same delay during config retrieval and given up.

The fix for this is simple:-

• Three-lines-icon -> Configuration -> Network Devices
• Tick the box alongside the device you're having trouble with (if multiple devices do them one at a time to avoid "credential profile" warnings), then click Edit (the pencil icon)
• Click Telnet/SSH and change the timeout from 60 seconds to a higher value (I picked 240)
• Click Update

Try a Configuration Backup again (at the above screen click the device name, Configuration Archive, Schedule Archive, Submit) and check the results in Administration -> Dashboards -> Job Dashboard.

[leitzf007]

This worked great, thank you! We weren't having problems with this timeout until recently. I wonder if there are certain settings on the wlc that cause delays during the display of the show run-config command.

[Eric R. Jones]

Hello, I have had this issue for some time. I get only when pushing templates from Configuration > Features & Technologies that I created. I reset the timeout to 240 and then 300 but it still failed with the same error.

We are running 3.1.7 update 02 with the 7.0 maintenance release and service packs 9, 10,11,15, and 16.

This error has always occurred but not sure why.

ej

[Jades Moreira]

Hi,

I had this same issue with three WLC (flex 7500 and 5508 models) and one Prime 3.1 (update 1.7 and Device Pack 15).

FAIL, Unsupported operation and Function run timed out 60000ms (60 sec)

To solve the problem was needed to change/edit the parameters of timeout from 60 to 300 in each WLC into Network Device, clicking "Update & Sync" and run Backup Job again.

Don´t forget to check credentials again to confirm the SSH/Telnet reachability.

[bill-anderson]

This also happens when trying to distribute software images with PI 3.1.2.