Prime Collaboration Deployment (PCD) – Dealing with Clusters

If we are planning to manager or if you planning to install new clusters this is the place to be. (Inventory –> Clusters)

Discovering an existing cluster.

Discovering and existing cluster is pretty straight forward. Go to clusters and click on the discover cluster. All the details are pretty self explanatory.

image

In the next step it will discover all the nodes which are associated with the publisher. It will take few minutes to recognize product and the product versions as per the below screen shot.

image

Next step which is cluster role assignment is quite important. PCD can be use to upgrade clusters in that case its important that PCD understand core functionality of specific node.  Its an optional step but still I highly recommend that. So in my case I have assigned all the functionalities of each node such primary call processor, primary TFTP etc.

image

Discovery status should be successful.

Adding a New UC cluster.

Personal favorite of PCD functions. I hate that blue screen. It’s a nightmare when you perform an installation  over slow connection. Key repetition issues, fat finger issue blah blah.

As of now PCD can not deploy OVA templates in to ESXi hosts. So basically you will need do it using vsphere client or web client. Like any other OVA. Once its deployed PCD can detect the VM and it can be use for UC application installation.

Navigate to clusters and click on “Define New UC Cluster”. It will walk you through the wizard.

Define a user friendly name for the new cluster.

image

In next step we will need to define the VM details such as

Hostname, IP details, Product functions for each UC application and finally select the VM template you just deployed. Just make sure that VM is power off. PCD will power it up and do the required installation based on the options which you configured during this step. In the backend it will mount a virtual floppy from the PCD NFS share. Answer file be generated based on the input of the add node page.

image

image

image

Once you define all the UC apps it will look like this.

image

My favorite part to eliminate fat finder mistakes and key repetition issues. Specify all the credentials and the certificate information.

image

During the wizard like any other UC application installation you will have options to configure DNS, NTP, MTU size and finally the time zone.

Once you complete the wizard you will notice there is a new cluster with cluster type defined as “New Install”.

image 

Well everything ready for the installation.

Uploading ISO files for the new installation.

You can use any SFTP client to transfer files to PCD. Logging with adminsftp user account with the password which you set for the OS administrator account. Once you log in select the fresh_install folder and transfer all the ISO images.

image

Installing the UC Application.

Navigate to task and click on the install. And go to add install task. It will automatically detect all the clusters with the cluster type set to “New Install”.

image

In next step we will need to select the ISO images which you transferred in the previous step. PCD is intelligent enough to understand CUCM and CUCM IM&P images.

image

You can schedule the installation. In my case im just gonna start it now.

image

You can specify the installation order. But in my case I have only 4 nodes. Anyway I need to wait till CUCM publisher and CUCM IM&P to be completed and then it will go ahead with the subscribers.

image

Finally review it and proceed with the installation. It will automatically start the VM and proceed with the installation. If you go to the settings of the VM you will notice there is a floppy image mounted to the VM from the NFS data store of the PCD.

 image

Once you click on the monitoring section you can see the status of your job.

image

You can click on the view log and you can get more details about the current progress. That’s it.

image

Hope this helps.

Leave a comment

Prime Collaboration Deployment (PCD) – Dealing with ESXi Hosts

This is the most interesting product I came across in recent past.

We had lot of unattended installation methods in the past but I was never a fan of it. But after CUCM IM and Presence became part of CUCM I was looking for good unattended installation methods. In those days we just install publishers of different UC apps with out interacting with each other.  But now in order to install CUCM IM and Presence publisher administrator needs to wait for the completion of CUCM publisher installation. And then you need to configure CUCM IM and Presence there with the domain details. But the PCD will do it behalf of the administrator.

PCD installation is quite straight forward. Nothing fancy easier than installing CUCM. Just the Cisco appliance operating system. I’m not going to discuss installation steps. Only difference between CUCM installation and PCD installation is that there is no application administrator. Basically your operating system administrator credentials will with the PCD GUI.

Discovering ESXi Hosts.

This step is required only when you are planning to install new UC application clusters in your vmware infrastructure. It will not interact with the vCenter instead of that it will directly access ESXi hosts.

Configuration steps –

Access PCD GUI and logging using OS credentials. Click on inventory and ESXi hosts. If you have any ESXi hosts added to the PCD it will be displayed here.

image

Click on the “Add ESXi Host” option to add a new ESXi hosts.

image

Pretty straight forward it should display its successful.

If you access your Storage configuration of ESXi you will notice there is a new data store named as hostname_NFS. This specific data store will be used in order to mount ISO images to VM’s and mount virtual floppy drives for unattended installation. Its a read only data store so what ever modifications needs to go through PCD GUI. Well more about it later.

Leave a comment

Session timeout in Cisco Communications Operating System

Aren’t we just hate when we need to login to CUCM again and again after session time outs ? mainly in lab systems. It is really really annoying. Same goes for CUCM CLI as well. Basically if you left your screen for 30 minutes you need to login again again. (Some times smoking breaks can drag for 31 minutes)

Default time is 30 minutes for both GUI and CLI.

View current session time out for CLI


admin:show cli session timeout

CLI session-timeout is set to 30 minutes for new CLI sessions.

View current session time out for GUI

admin:show webapp session timeout

The current session-timeout used for web sessions and applications is 30 minutes.

As per below command im setting the CLI session time out for 99999 minutes. Value range is from 5 minutes to 99999 minutes.

admin:set cli session timeout 99999

Continuing with this operation will change the session timeout for all new CLI sessions to 99999 minutes.
Continue (y/n)?y
CLI session-timeout updated to 99999 minutes.  You must restart your session or establish a new CLI session for the update to take effect.
CLI session-timeout is set to 99999 minutes for new CLI sessions.


As per below command im setting the GUI session time out for 35000 minutes. Value range is from 5 minutes to 35000 minutes. But after executing the below command it will force you to restart the Cisco Tomcat service.

admin:set webapp session timeout 35000

Continuing with this operation will set the session-timeout for web sessions to 35000 minutes after the ‘Cisco Tomcat’ service has been restarted or after the server has been rebooted.
Continue (y/n)?y
Tomcat session-timeout updated to 35000 minutes.

The ‘Cisco Tomcat’ service needs to be restarted for the changes to take effect immediately.  This will disconnect active web sessions.
Continue (y/n)?y
Don’t press Ctrl-c while the service is getting RESTARTED.If Service has not Restarted Properly, execute the same Command Again
Service Manager is running
Cisco Tomcat[STOPPING]
Cisco Tomcat[STOPPING]
Cisco Tomcat[STARTING]
Cisco Tomcat[STARTING]
Cisco Tomcat[STARTED]

The ‘Cisco Tomcat’ service was restarted successfully.  New web sessions will timeout after 35000 minutes.

The current session-timeout used for web sessions and applications is 35000 minutes.

Leave a comment

Shared Virtual Machines in vmware workstation.

Well it was there for few editions so far. Never had a requirement for me to test it. Since I moved away from hyper-V completely I decided to give it a shot. And this is what I have notice. 

Well first things first. what is a shared VM. Well term it self miss guides me. We don’t have shared SAN’s or management tools to manager vmware workstation. When you share VM’s across multiple workstations. Basically its only to interact with the VM. In another words remote KVM. Reason why I found this interesting because I have 5 VM’s running windows 7 with Jabber software. I just wanted monitor VM behavior from single point of view. But the bad news is you cant keep those VM’s across multiple hard drives with in the same vmware workstation hosts. Yes you cant.Your shared VM’s will reside in one folder. and you can configure it from the Edit –> Preferences –> Shared VMs

Time for the configuration steps.

Configure the file path for your VM’s. Default it will store everything in C drive. So move it some where else. Only frustration is I can have only one folder for shared VM’s. So don’t think about sharing all the VM’s that you have with in the host. Validate which VM you really need to share. For example if you are running a domain controller you can control it well with the remote desktop utilities. If you are interacting with the VM using sound and webcam (which is not so great with remote desktop) this is the best option for you. Make sure you create a folder inside the partition and  select that folder as the shared vm folder. Don’t select the whole drive. It will make it unreadable from the windows. Yes it happened to me. I couldn’t find any details about it. I had to reformat it and use it. Well after having VM’s registered as the shared VM’s it will be nightmare to change the file path. You will need to remove all the shared VM’s and then change the file path. So select it wisely. And you will be stuck with the file path for sometime.

image

If you have already create the VM just make sure you power it off. and Drag and drop to the shared VMs Icon or you can right click on the VM –> manage –> click on share.

image

Then it will move the VM to the shared folder and register it again with the vmware workstation. image

Successfully we have added a shared VM. So your VM should be listed under shared VM section in the vmware workstation snapping. Next is to add remote hosts in to this VMware workstation. I have already shared few VM’s there. Click on the file and select connect to server option. It will prompt you to add details for the remote host.

Server name – IP address of the remote host which runs the vmware workstation.

User name – windows credentials of that remote host. In my case its just the local administrator account in windows 2012. and of course the password.

image

Once it successfully connected You can see the VM’s and some statistics related to your remote host.

image

Once I select one of those VM’s which are listed under remote host, it will open it up in separate tab in vmware workstation.

Single point of control. Hope this post will help someone. 

Winking smile

Leave a comment

CUCM SIP Trunk options ping.

Well we all are use to this luxury of pinging to remote end and get the health status of the remote peer. But when its come to SIP, ICMP is not that ideal because it validates only layer 3 status. We have a similar utility in in most of the sip related devices to check the status of remote sip peer. Basically it will provide you layer 7 health status.

But this was not available in Cisco unified communications manager. I notice this in the CUCM version 10. But never had time to test it. But in version 10.5 under SIP trunk configurations I noticed that it has separate section call “SIP Trunk status” under statue it complained that service status is unknown and option ping is not available. Quite a handy feedback related to your configuration.

image

In order to enable option ping you will need to edit the sip profile which is associated with the SIP trunk. You can access the sip profile via Device –> Device Settings –> Sip Profile.

I always like to leave default sip profile as it is because there are many devices associated with that. I have the flexibility to copy that specific sip profile and edit it accordingly. here are the options related to the SIP options ping.

image

Just enable the Option ping to monitor destination option and save it. Go to the sip trunk configuration and change the sip profile which you configured with SIP option ping. and reset the trunk. Bingo it will let you know the service status and for how long it was running.

image

Well this can be the first troubleshooting step that you can perform when your users complain about PSTN outages

By the way happy new year.

Winking smile

Leave a comment

Hyper-V or VMware workstation ?

Well i land here with 3 laptops and now i have 7 Desktop PC’s and 4 Laptops. (And some of you ask where is my salary) Well Since i started working from home i always preferred to have a separate lab for my self and as a tech enthusiast i always want to do all the crazy experiments with out braking some configuration belong to some one else. 

Anyway coming back to the topic i have 5 hosts running vmware esxi and 2 hosts running windows 2012 R2 with hyper-v role. Well first of all about the memory utilisation and performance i didn’t notice anything. Its pretty much the same. Since all the Cisco UC applications are not working with hyper-v i decided to use vmware environment for the Cisco UC applications and Hyper-v machines for the active directories and windows 7 PC’s to run soft phones such as jabber and CIPC. 

But recently i had a crazy idea about testing some things with jabber and video. So first things first i thought of buying few webcams. After thinking about my LYNC deployment and CUCM IM i decided to go ahead with mid range cameras. So i end up buying logitech C270. Well its not fancy but it is good enough for the task. So i end up buying 6 web cams. For each hyper-v host will have 3 cameras. So each VM will have one cam for each. Since i was running windows 8 VM’s for testing i never thought how audio passthrough really works between the hyper-v host and the Virtual machine. Well it was working fine because of the hyper-v enhanced session mode. well as long as it was working fine i never bothered. But now I need to passthrough video traffic from hyper-v host to the virtual machine. So there were lot of examples and configuration guide lines in different blogs regarding this but no one didn’t talk about this specifically. But in one of those blogs i saw one nikon dslr cam was detected. And in the mean time i saw lot of comments regarding plugging an iPhone to the VM. I never tried it but they claimed it was never detected by the virtual machine.

Anyhow i tried my best to use enhanced session mode for this task but it never detected my webcam. i was able to uplink my printer and an usb drive. But again that usb drive was detected by the VM as a network share from hyper-v host. So bottom line is i was unable to use hyper-v for my testing because enhanced session mode never detected my camera correctly. I thought its something to do with the host operating system. But webcam was detected correctly by the hyper-v host. In one of my laptops i activated hyper-v service  with windows 8.1 and i was getting the same result.

No need to say how unhappy i was and i had to switch from hyper-v to vmware workstation because of that. I was sure that USB passthrough really works in this sort of cases than hyper enhanced session mode. My guess was correct and now i have full vmware lab for testing.

Bye bye hyper-v 🙂   

 

Leave a comment

Another Christmas.

Another Christmas.
Another Year.
Another prayer at the church.
Another occasion for a drink.
Another whisky bottle at the table.
Another animal goes to the table.
Another bird goes to the table.
Another fresh start for some.
Another mourning for some.
Another feast for the rich.
Another licking for the poor.
Merry Christmas ho ho ho !!!

Leave a comment