Is this overkill ?

I am planning on building a new server to have multiple call managers running on VM (hq/br1/br2/etc).. mainly b/c my current sever cant support more than 8 gigs of ram. I plan on going with 16.

I have skimmed through the cipt2 material and it doesn't really go into running multiple servers..

what i have noticed is the book goes through 1 CM server with a bunch of remote offices running CME.. I only saw a few examples where Site 1 and Site 2 had their own clusters.

so my question is.. do i really need to build a dedicated server to run 4-7 Call managers (pub/subs) in order to really be prepared ?

my current server has 8 gigs and currently has 1 CM7 VM w/2 gigs allocated. I can certainly host another CM7/8 and still have enough ram left. CPU utilization wont be a issue due the lack of demand on the server (LAB).

your thoughts are welcome.

Thank you.

Find more posts tagged with

Comments

chmorin

IMHO you really only need on site to have a PUB/SUB combo to get a grasp of clustering. CIPT2 does go a little bit into having multi-site cluster implementation and how you can route between them. Clustering over a VM has its issues, but for a LAB you're probably okay.

To have your lab scale the way you probably want to, (and to keep up with the future complexities that will come into play) if you can afford to, double your RAM. Virtualizing CUCM servers is one thing, but you can also consolidated GNS3 and host VM's onto the same box. Trust me, you could find a use for all 16 GB if you sat down and made a really complex lab out of it (especially in voice; think of the CIPC's on hosts).

If I could build a server right now (if my finances were actually supporting the work I do, but that's another rant for another day) I would start with 16GB. But that's just me.

tokhss

I am not sure i need GNS3 right now?

my lab consists of 1x server (8giger).. 3x 2800s, 1x 2650xm, POE switches, and multiple Phones.

what i am going back n forth (with myself) about is running the 8 giger a bit longer.. running 2 call managers on it to test site 1 -CM , site 2 -CM and the rest of the sites with CME.

or... if the material really calls for it to run a new server (16gig) aand host 4-6 call managers on it.

thats the part i am not sure about.... Will the material call for a new server ? or will i be able to do what i need to do right now with 2 call managers?

IPDirect

Guess it really comes down to a cost vs benefits question, but only you can answer that. I know for me, if I can lab/understand the fundamentals, I'm basically good to go without labbing every possible senario.

Even the CCIE lab only requires a single PUB+SUB cluster to prepare...just sayin

tokhss

LOL.. really ?

chmorin

IPDirect wrote: »

Even the CCIE lab only requires a single PUB+SUB cluster to prepare...just sayin

I agree. If you are teetering on the cluster thing, you really only need to do it once to understand how its done. Don't spend extra money if you have the physical hardware to separate the resources from your server.

tokhss

Thats really what it comes down to for me.. I needed to know if i really needed like all those servers to represent 4-5 locations all at once on the network.

if all i need are a couple CM's in VM to test pub/sub and test site1 site2 then **** im gonna save the build money and put it towards other hardware. lol

pitviper

For the purpose of CCVP (Err, CCNP-Voice) it’s overkill. I just fired up a couple of CUCM VMs on my laptop and bam – cluster 2!

For the CCIE you’ll probably want a second server (processing power / disk usage is the issue here, not RAM) – but not for clustering CUCM , more so for running CUCM-PUB, CUCM-SUB, Unity Connections, Presence, and UCCX, all concurrently.

chmorin

pitviper wrote: »

For the purpose of CCVP (Err, CCNP-Voice) it’s overkill. I just fired up a couple of CUCM VMs on my laptop and bam – cluster 2!

For the CCIE you’ll probably want a second server (processing power / disk usage is the issue here, not RAM) – but not for clustering CUCM , more so for running CUCM-PUB, CUCM-SUB, Unity Connections, Presence, and UCCX, all concurrently.

Random question. CUPS - Any learning material that you would suggest?

pitviper

The CUCM SRND Presence sections give a nice overview for starters. For configuration, I’ve been using the “Configuration and Maintenance” guide from Cisco.com:

Cisco Unified Presence, Release 7.x -- Configuration and Maintenance of Cisco Unified Presence - DocWiki

Not a whole lot of 3rd party material that I’ve come across. The CCBOOTCAMP CCIE Voice Technology Lab Workbook has a couple of nice integration labs, but that's about all that I’ve found.

jahsoul

I've been skimming over Ch. 23 in the CUCM 8.x SRND and just follow the links that they give. Not saying that I'm ready to learn it but I was just trying to see if it was covered.

tokhss

well, since I will begin my CCIE track right after i obtain my CCVP, i might as well get a 2nd server. with everything going around the world these days. i have feeling prices are going to go up on components.

additionally.. for those curious like me.. here is a link to the Cisco Wiki on CCM8 + Applications for Virtualization

Unified Communications Virtualization - DocWiki

shodown

I run a PUB/SUB, UCCX, UC, and CUPS on a HP prolient with around 6 gigs of ram you don't need much as much as you think you will need. You will also not need Multiple clusters for CCIE V. You will need routers with enough resources to run the correct IOS(you have 28XX those should do everything) 3750 or 3560 Switches and a Etherswitch module to test everything out. You can skimp on some of it and get rack rentals, but make sure you homelab can do overall 75-85 percent of what you need to do and run rack rentals for the rest of it. I'm planning on using left Over CCNP items (3725's/2651XM) to handle the majority of the task and getting 1 3750 and Eitherswitch module to run QOS and making up everything else on rack rentals to keep cost down.

tokhss

Shodown.. if you're running multiple clusters.. would that not warrant a beefy server? I dont plan on rack renting and really want to complete my home setup as much as i can ..cost is not much of a concern right now.. its more of an investment to me.

my 2800s are running max mem/flash.. i do have one 3750-48ps. Now back to running multiple clusters.. this was my initial concern. Multiple sites (clusters) .. lets say 3 sites for example .. would 6 gigs of ram cut it? what do you recommend ?

shodown

tokhss wrote: »

Shodown.. if you're running multiple clusters.. would that not warrant a beefy server? I dont plan on rack renting and really want to complete my home setup as much as i can ..cost is not much of a concern right now.. its more of an investment to me.

my 2800s are running max mem/flash.. i do have one 3750-48ps. Now back to running multiple clusters.. this was my initial concern. Multiple sites (clusters) .. lets say 3 sites for example .. would 6 gigs of ram cut it? what do you recommend ?

I currently don't have experience with multicluster setups in a lab. If I had 3 sites I wouldn't start looking at them for there own clusters until they moved beyond the requirements of SRST and CME at there site, or they had Call Center on site. So far in my VOIP career I've had less than 10 customers who met that requirement. I would focus on what you will be tested on for right now and you can configure ICT's and so on in call manager you just won't have a way to prove they work. You can setup CME with one of your 2811's and test out Gatekeeper and so on and get a good idea of how multiple sites interact with each other. The other options is to build you another sever like it was mentioned below, but you dont' want to spend a bunch of money and not get any return value on it. You already got a pretty bad ass lab. Exploit it to its limits before you decide you need to add on to it.

tokhss

Ok...so i just took a peek at the CCIE-V lab @ Cisco. Seems like i was getting out of hand now that i see whats listed below. Seems like i might add a couple more devices to my list but thats really gonna be about it. feeling a bit relieved

CCIE Voice Lab v3.0 Equipment and Software List
Passing the CCIE Voice Lab Exam requires a depth of understanding difficult to obtain without hands-on experience. Early in your preparation, you should arrange access to the equipment listed below:

Lab Equipment:
Cisco MCS-7845 Media Convergence Servers << VM Ware
Cisco 3825 Series Integrated Services Routers (ISR) << sub with 2811
Cisco 2821 Series Integrated Services Routers (ISR) << sub with 2811
ISR Modules and Interface Cards+ VWIC2-1MFT-T1/E1 << Check!
- PVDM2 << Check!
- HWIC-4ESW-POE << gotta buy

- NME-CUE << uc520w prior experience. Will prob ebay the NM soon.
Cisco Catalyst 3750 Series Switches << Check!
IP Phones and Soft Clients << Check!

Software Versions
Any major software release which has been generally available for six months is eligible for testing in the CCIE Voice Lab Exam.
Cisco Unified Communications Manager 7.0 << Check!
Cisco Unified Communications Manager Express 7.0 << Check!
Cisco Unified Contact Center Express 7.0 << gotta find...
Cisco Unified Presence 7.0 << gotta find...
Cisco Unity Connection 7.0 << gotta find...

All routers use IOS version 12.4T Train. << Check!
Cisco Catalyst 3750 Series Switches uses 12.2 Main Train << Check!

Network Interfaces
Fast Ethernet << Check!
Frame Relay << Check!

Telephony Interfaces
T1 << Check!
E1 << Check!

tokhss

1.0 Implementing and Troubleshooting Campus Infrastructure and Services
1.1. VLANs
1.1.1. Voice VLANs
1.1.2. Data VLANs
1.2. DHCP
1.2.1. Cisco Unified Communications Manager Server-Based DHCP
1.2.2. Cisco IOS DHCP
1.3. TFTP
1.3.1. TFTP as a Cisco Unified Communications Manager Service
1.3.2. TFTP Options
1.3.3. TFTP Redundancy
1.4. NTP
1.4.1. Cisco IOS NTP
1.4.2. Cisco Unified Communications Manager NTP
2.0 Implementing and Troubleshooting Cisco Unified Communications Manager Endpoints
2.1. Cisco Unified Communications Manager SCCP Endpoints
2.1.1. Device Registration
2.1.2. Device Template
2.1.3. Softkey Template
2.1.4. Device-Related Profiles and Information
2.1.5. Protocol-Specific Profiles and Information
2.1.6. Directory Number and Related Information
2.2. Cisco Unified Communications Manager SIP Endpoints
2.2.1. Device Registration
2.2.2. Device Template
2.2.3. Softkey Template
2.2.4. Device-Related Profiles and Information
2.2.5. Protocol-Specific Profiles and Information
2.2.6. Directory Number and Related Information
3.0 Implementing and Troubleshooting Cisco Unified Communications Manager Express Endpoints
3.1. Cisco Unified Communications Manager Express SCCP Endpoints
3.1.1. Device Registration
3.1.2. Ephone Template
3.1.3. Directory Number and Related Information
3.1.4. Ephone Features
3.2. Cisco Unified Communications Manager Express SIP Endpoints
3.2.1. Device Registration
3.2.2. Voice Register Template
3.2.3. Directory Number and Related Information
3.2.4. SIP Phone Features
4.0 Implementing and Troubleshooting Voice Gateways
4.1. T1/E1 PRI
4.1.1. Layer 1 Configuration and Options
4.1.2. Layer 2 Configuration and Options
4.2. T1/E1 CAS
4.2.1. Layer 1 Configuration and Options
4.2.2. Layer 2 Configuration and Options
4.3. H.323
4.3.1. Cisco Unified Communications Manager H.323 Configuration and Options
4.3.2. Cisco IOS H.323 Configuration and Options
4.4. MGCP
4.4.1. Cisco Unified Communications Manager MGCP Configuration and Options
4.4.2. Cisco IOS MGCP Configuration and Options
4.5. SIP
4.5.1. Cisco Unified Communications Manager SIP Configuration and Options
4.5.2. Cisco IOS SIP Configuration and Options
4.6. H.323 RAS (Gatekeeper)
4.6.1. Cisco Unified Communications Manager Gatekeeper Configuration and Options
4.6.2. Cisco IOS Gatekeeper Configuration and Options
4.6.3. Cisco IOS Gateway Configuration and Options
4.7. Cisco Unified Border Element (IP-IP Gateway)
4.7.1. Cisco IOS VoIP Protocol Configuration and Options
5.0 Implementing and Troubleshooting Call Routing Policies
5.1. Route Patterns and Dial Peers
5.1.1. Cisco Unified Communications Manager:
5.1.1.1. Route Pattern
5.1.1.2. Route Group
5.1.1.3. Route List
5.1.1.4. Hunt Pilot
5.1.1.5. Hunt Group
5.1.1.6. Hunt List
5.1.2. Cisco IOS Gateway:
5.1.2.1. Dial Peers
5.2. Digit Manipulations and Translations
5.2.1. Cisco Unified Communications Manager:
5.2.1.1. Translation Pattern
5.2.1.2. Transformation Pattern
5.2.2. Cisco IOS Gateway:
5.2.2.1. Voice Translation Rules and Profiles
5.2.2.2. Number Expansion
5.2.2.3. Dial-Peer Digit Manipulations
5.3. Class of Service
5.3.1. Cisco Unified Communications Manager:
5.3.1.1. Access List
5.3.1.2. Time Period
5.3.1.3. Time Schedule
5.3.1.4. Partition
5.3.1.5. Calling Search Space
5.3.1.6. Client Matter Codes
5.3.1.7. Forced Authorization Codes
5.3.2. Cisco IOS Gateway:
5.3.2.1. Class of Restriction
5.4. Route Selection Preference and Redundancy
5.4.1. Cisco Unified Communications Manager:
5.4.1.1. Route Pattern
5.4.1.2. Route Group
5.4.1.3. Route List
5.4.2. Cisco IOS Gateway:
5.4.2.1. Dial-Peer Preference
5.5. Mobility and Single-Number Reach
6.0 Implementing and Troubleshooting High-Availability Features
6.1. Cisco Unified SRST
6.1.1. Cisco Unified Communications Manager SRST Configuration and Options
6.1.2. Cisco IOS SRST Configuration and Options
6.2. AAR
6.2.1. Cisco Unified Communications Manager AAR Configuration and Options
6.3. Call Forward Unregistered (CFUR)
6.3.1. Cisco Unified Communications Manager CFUR Configuration and Options
7.0 Implementing and Troubleshooting Media Resources
7.1. Codec Selection and Flexibility
7.1.1. Endpoint Codec Configuration Options
7.1.2. Cisco IOS Gateway Codec Configuration Options
7.1.3. Codec Used by Other Applications (e.g., MOH)
7.2. Conference Bridge (CFB)
7.2.1. Cisco Unified Communications Manager Software Conference Bridge
7.2.2. Cisco IOS Hardware Conference Bridge
7.3. Transcoder
7.3.1. Cisco Unified Communications Manager Software MTP
7.3.2. Cisco IOS Hardware Transcoder
7.4. Music on Hold (MOH)
7.4.1. Cisco Unified Communications Manager Music on Hold Server Configuration and Options
7.4.2. Unicast and Multicast MOH
7.4.3. Cisco IOS MOH
7.5. Media Resources Preference and Redundancy
7.5.1. Media Resource Group
7.5.2. Media Resource Group List
7.6. Other Cisco Unified Communications Manager Media Resources
7.6.1. Annunciator
8.0 Implementing and Troubleshooting Supplementary Services
8.1. Call Park
8.1.1. Cisco Unified Communications Manager Call Park
8.1.2. Cisco Unified Communications Manager Express Call Park
8.2. Call Pickup
8.2.1. Cisco Unified Communications Manager Call Pickup
8.2.2. Cisco Unified Communications Manager Express Call Pickup
8.3. Barge
8.3.1. Cisco Unified Communications Manager Barge
8.3.2. Cisco Unified Communications Manager cBarge
8.4. Callback
8.5. Other Supplementary Services
8.5.1. Cisco Unified Communications Manager Intercom
8.5.2. Cisco Unified Communications Manager Express Intercom
9.0 Implementing and Troubleshooting Other Cisco Unified Communications Manager Voice Applications
9.1. Extension Mobility
9.1.1. Cisco Unified Communications Manager Extension Mobility Configuration and Options
9.2. Cisco IPMA
9.2.1. Cisco IPMA Configuration and Options
9.3. Others
9.3.1. Attendant Console Configuration and Options
10.0 Implementing and Troubleshooting QoS and CAC
10.1. Layer 2 and Layer 3 Traffic Classifications and Policing
10.1.1. CoS Marking
10.1.2. IP Precedence Marking
10.1.3. DSCP Marking
10.1.4. Policing
10.1.5. Traffic Shaping
10.2. Layer 2 and Layer 3 Queuing Mechanisms
10.2.1. Low Latency Queuing
10.2.2. Priority Queuing
10.2.3. Shaped Round-Robin Queuing
10.3. Layer 2 LFI
10.3.1. Header Compression
10.3.2. Frame Relay Link Fragmentation and Interleaving
10.4. RSVP
10.5. Call Admission Control
11.0 Implementing and Troubleshooting Messaging
11.1. Cisco Unity Connection
11.1.1. System Integration Options
11.1.2. Subscriber Configuration Options
11.2. Cisco Unity Express
11.2.1. System Integration Options
11.2.2. Subscriber Configuration Options
11.3. Call Handling and Routing
11.3.1. Call Handling and Routing Configuration and Options
12.0 Implementing and Troubleshooting Cisco Unified Contact Center Express
12.1. Advanced Configuration
12.1.1. Application Management
12.1.2. Script Management
12.1.3. Prompt Management
12.2. Script Customization
12.3. Redundancy
12.3.1. CTI Redundancy
13.0 Implementing and Troubleshooting Cisco Unified Presence
13.1. Cisco Unified Communications Manager Presence
13.1.1. BLF Configuration and Options
13.2. Cisco Unified Presence Server Integration
13.2.1. Cisco Unified Presence Server and Cisco Unified Communications Manager Integration Options
13.2.2. Cisco Unified Personal Communicator Configuration Options
13.2.3. Other Cisco Unified Presence Features

IPDirect

tokhss wrote: »

Ok...so i just took a peek at the CCIE-V lab @ Cisco. Seems like i was getting out of hand now that i see whats listed below. Seems like i might add a couple more devices to my list but thats really gonna be about it. feeling a bit relieved

Haha, you were getting a bit crazy. You may also want to check out the rack rentals topologies at ipexpert/internetwork expert for ideas on your lab hardware. Off the top of my head I believe:

IPX
---
HQ:
2811
3750
Vmware cluster
vwic-2mft-t1
pvdm2-32

BR1:
2811
HWIC-4ESW
vwic-2mft-t1
pvdm2-16

BR2:
2811
HWIC-4ESW
vwic-2mft-e1
CME installed
AIM-CUE
pvdm2-16

PSTN/WAN
2811
2-vwic-2mft-t1
1-vwic-2mft-e1

INE:
---
HQ:
2811
3750
VMware cluster
vwic-2mft-t1
pvdm2-32

BR1:
2811
HWIC-4ESW
vwic-2mft-t1
pvdm2-16

BR2:
2811
3550
vwic-2mft-e1
CME installed
NM-CUE
pvdm2-16

PSTN/WAN
3725
NM-HDV w/ 2 pvdm-8s
2-vwic-2mft-t1
1-vwic-2mft-e1

Might need to double check all the info but it should be damn close. I recently started building out my CCNP Voice lab off this model because I know I'll be launching into the CCIE right afterwards.

tokhss

you know.. my lab is very close to both of those.. I was actually looking at the INE labs right now.

Honestly, I think you can get by with a single 3750PS and just vlan / segment the offices rather than getting a HWIC POE for the branches. Either way.. for the sake of learning and most importantly.. buying toyz! I plan on getting it..

what i also noticed that the NME-CUE now replaces the NM-CUE in the v3 lab. same ish.. v7.0 replaces v3.3 or something? prob going to get an NM-CUE also. that should for the most part complete my lab .. only thing left is obtaining the UCCX, and CUP SW.

phew.. here i was about to build a monster machine (itch) really lol.

IPDirect

tokhss wrote: »

and most importantly.. buying toyz! I plan on getting it..

Tell me about it...if I spent more time learning the material and less time buying/researching hardware, i'd prolly be further along in my CCNP Voice than I am. Oh well, all part of the fun!

tokhss

I think in our area of the tech world, you really gotta love doing this stuff.. calling hardware toys is part of the mental state of mind and lifestyle

pitviper

tokhss wrote: »

you know.. my lab is very close to both of those..

That’s because you used good reference material

tokhss wrote: »

Honestly, I think you can get by with a single 3750PS and just vlan / segment the offices rather than getting a HWIC POE for the branches.

You are correct - the bad thing about the HWIC and NME switch modules is that in order to use PoE, you also have to upgrade the router PSU.

tokhss

pitviper wrote: »

That’s because you used good reference material

You are correct - the bad thing about the HWIC and NME switch modules is that in order to use PoE, you also have to upgrade the router PSU.

LOL.. everybody.. Pits setup was my inspiration!! .. and its true.. i did ref your design along side others

totally forgot about the PSU upgrade.. oh well.. clicking on ebay now lol

damn.. PSU is $$

tokhss

so the PSU is 200 new .. noticed a psu cable. for 4 bucks or something.. i take it thats needed also ?

pitviper

tokhss wrote: »

so the PSU is 200 new .. noticed a psu cable. for 4 bucks or something.. i take it thats needed also ?

Yeah, you need the cable. They typically come with a cable when you purchase them new/direct. The used ones may or may not. Be careful with the “open box” ones – Have the seller check the part number on the PSU. A lot of times (and I’ve done this myself) the inline power unit is used, and the standard PSU put back into the box. Then a liquidator comes along and goes by the part number on the box!

tokhss

after consideration.. i decided to not even touch the psu upgrade path for the hwic.. nor will i get the unity express module right now..

200 for psu new + cable + hwic = 400+ and aim-cue 300-400 .. possible total of 700-800 .. not feeling that right now LOL...

i got some other ideas in mind that may better assist with my studies..

well, i just got bit by the bugg...

just picked up:

3745 256/128
NM-1FE-2W 2x WIC-1DSU-T1 (cuz i never setup data t1)
NM-HDV + 3 pvdm12's

converting my 2650xm into a GK. moving my 2811 pstn/frame into a location and setting up the 3745 into the frame/pstn.

gotta love ebay.. all for less than $320

pitviper

tokhss wrote: »

well, i just got bit by the bugg...

Hehehe, just add this to your local hosts file:

174.133.7.131 ebay.com

problem solved

tokhss

LOL.... good one.

GuyManDude

Save your money and just buy the power supplies for the phones instead of running them off the switch poe. The phone power supplies are less than $10.00 ea off Ebay...

tokhss

GuyManDude wrote: »

Save your money and just buy the power supplies for the phones instead of running them off the switch poe. The phone power supplies are less than $10.00 ea off Ebay...

I dont lab everyday all day.. i get to lab @ home 1-3 days per week and thats when im really home lol. my energy bill is 50 bucks per month.

right now im doing a lot of sip testing so im not really doing a lot with my phones physically. i am routing through a soft client like X-lite .. im gonna redo my network this week and setup a couple sip phones / profiles along side some skinny phones.

at some point, my energy bill will go up but even then. not by much.. and honestly, those bricks are damn ugly, more wires, and more outlets .. meh.

IPDirect

tokhss wrote: »

200 for psu new + cable + hwic = 400

I'm on the same boat as you.

I'm wondering if getting a regular hwic-4esw (non-poe) and wall plugs for two of the phones would suffice. Cost would be about $150 vs $400, but I'm not familiar with the extent of the role the hwic-4esw-poes play for the ccie.