I was up at 2:30AM this morning for a Cisco 7600 install. So, I have not gotten much completed on the SNAT programming. So far, I have the SNMP poller code able to do a single MIB poll (sysObjectID for example) and I am currently working on the getBulk and getNext functions. Once I have those completed, the poller will be fairly well finished.
I have VMware working on my laptop so I will be porting the rest of the application (besideds the poller and collector) to the virtual appliance and see how those go.
I was also able to get the Wiki going with some more information. It took me a few minutes to get everything working correctly, but now it all seems to be working. Check out the lab topology at http://dragool.net/wiki/LAB.
Time to head to bed now.
Wednesday, May 27, 2009
Monday, May 25, 2009
Lab Documentation Completed
I have finally completed the final set of lab documenation for all of the equipment in the test lab. I am finishing the Wiki post with all of the basic configurations for the equipment and and the detailed documentation on the configuration files.
Below is the updated diagram:
In addition, I think I am going to change the way the information is displayed on the screen for the interface data so the graphs are not shown on the large list of interfaces until after the interface is select. This should allow a better use of the CPU resources on the AXP platform (especially for 6500 type systems with hundreds of interfaces).
I am also working on the diagramming process again to see if I can make it more scalable for larger networks. I will put more information on that process later.
Below is the updated diagram:
With all of the routers working correctly again (and everything reachable), I am not working on the porting of SNAT To the AXP platform on the virtual machine environment. I think I am going to have issues getting all of the Perl modules moved over to the platform (primarily because I had a huge issue getting them working on full system).
My current plan involves moving some of the more complicated functions to java applications (JAR files) that have all of the modules integrated. These currently involve the following:
- SNMP Performance Polling
- Telnet/SSH/SNMP information collection
- Performance Data Graphs
In addition, I think I am going to change the way the information is displayed on the screen for the interface data so the graphs are not shown on the large list of interfaces until after the interface is select. This should allow a better use of the CPU resources on the AXP platform (especially for 6500 type systems with hundreds of interfaces).
I am also working on the diagramming process again to see if I can make it more scalable for larger networks. I will put more information on that process later.
Sunday, May 24, 2009
Cisco AXP Developer Contest
I should've posted this earlier, but I am one of the ten finalist in the Cisco AXP Developer Contest http://www.cisco.com/web/solutions/axpdev/index.html.
Team SNAT is my team (I am the only member). I am very surprised by the fact that I got in as a finalist because of the number of entries (over 900). This is the primary reason I am reconfiguring the lab. In order to get the development working, I need to not only have the network lab enabled and running, but also have a VM environment simulating the AXP.
So far, I have the lab up and running, but no VM environment yet. I plan on getting the VMware server running on my laptop since I currently only have a single server for the lab. If that does not work well enough, I may just end up buying another server to act as my VMware host. I just need to find one that supports more than 8 GB of RAM (so I can do a full Microsoft environment for the OCS as well).
Team SNAT is my team (I am the only member). I am very surprised by the fact that I got in as a finalist because of the number of entries (over 900). This is the primary reason I am reconfiguring the lab. In order to get the development working, I need to not only have the network lab enabled and running, but also have a VM environment simulating the AXP.
So far, I have the lab up and running, but no VM environment yet. I plan on getting the VMware server running on my laptop since I currently only have a single server for the lab. If that does not work well enough, I may just end up buying another server to act as my VMware host. I just need to find one that supports more than 8 GB of RAM (so I can do a full Microsoft environment for the OCS as well).
Wednesday, May 20, 2009
Lab mostly working
After spending a little time last night, I finally got the lab working. I was having some strange issues getting the ATM portion of the lab working (I was getting some errors about being unable to simulate the CPU stack or something). This would cause some random hangs in my 7200 routers acting as the service provider (in the MPLS cloud) to the ATM network. To correct this, I just removed the ATM components of the lab for now.
As of right now, I was able to get the entire lab up and running (minus the ATM components of course) and I have about half of the lab gear with a basic configuration. The MPLS network is about half finished (just need to finsish the BGP configuration) and then get the rest of the frame relay network completed.
In order to get the lab working, I have created five dynamips instances. Although I am using the 64-bit version of the code, I belive the multiple instances allows for the better CPU utilization as I see all of the CPUs being utilized. Here is what I use for the 'startlab.sh' file that starts the lab:
I use the hypervisors from 7200 to 7204 to split the load across the CPUs. I had a difficult time getting them all going and stable because they were originally all trying to write to the same directories. In order to get around this, I used the seperate directories for each of the hypervisors (along with seperate log files).
I will try to get the configurations and the updated detailed diagram up soon.
As of right now, I was able to get the entire lab up and running (minus the ATM components of course) and I have about half of the lab gear with a basic configuration. The MPLS network is about half finished (just need to finsish the BGP configuration) and then get the rest of the frame relay network completed.
In order to get the lab working, I have created five dynamips instances. Although I am using the 64-bit version of the code, I belive the multiple instances allows for the better CPU utilization as I see all of the CPUs being utilized. Here is what I use for the 'startlab.sh' file that starts the lab:
#!/bin/sh
# Make the directory structures
mkdir -p /usr/local/src/lab
mkdir -p /usr/local/src/lab/7200
mkdir -p /usr/local/src/lab/7201
mkdir -p /usr/local/src/lab/7202
mkdir -p /usr/local/src/lab/7203
mkdir -p /usr/local/src/lab/7204
# Starting dynamips instances
cd /usr/local/src/lab/7200
/usr/local/bin/dynamips -H 7200 > /usr/local/src/lab/7200.log.txt &
cd /usr/local/src/lab/7201
/usr/local/bin/dynamips -H 7201 > /usr/local/src/lab/7201.log.txt &
cd /usr/local/src/lab/7202
/usr/local/bin/dynamips -H 7202 > /usr/local/src/lab/7202.log.txt &
cd /usr/local/src/lab/7203
/usr/local/bin/dynamips -H 7203 > /usr/local/src/lab/7203.log.txt &
cd /usr/local/src/lab/7204
/usr/local/bin/dynamips -H 7204 > /usr/local/src/lab/7204.log.txt &
I use the hypervisors from 7200 to 7204 to split the load across the CPUs. I had a difficult time getting them all going and stable because they were originally all trying to write to the same directories. In order to get around this, I used the seperate directories for each of the hypervisors (along with seperate log files).
I will try to get the configurations and the updated detailed diagram up soon.
Tuesday, May 19, 2009
Daily Digest for May 19th
 | Heading to gym. [#] |
 | Published Daily Digest for May 19th |
| Implementing Cisco VPLS today. [#] |
| Published Setting Up the New Lab Server |
Setting Up the New Lab Server
I got done with a client engagement early today (before lunch), so I decided to take the time to setup the lab server. The server is running on a white box home-built server. The server has the following specification:
The architecure of the layout of the network is shown below:
[caption id="attachment_26" align="aligncenter" width="600" caption="SNAT Network Toplogy Diagram"]
[/caption]
The MPLS, ATM and Frame Relay clouds are actually routers (hidden from the topology) allowing for integration of the different network topologies.
After I get all of the functions operating correctly on the server, I will post the topology file.
- 8 Gigabytes of RAM
- Intel Quad Core 2.4GHz CPU
- 2 - 500 GB SATA hard drives
- Dual network interfaces
The architecure of the layout of the network is shown below:
[caption id="attachment_26" align="aligncenter" width="600" caption="SNAT Network Toplogy Diagram"]
[/caption]The MPLS, ATM and Frame Relay clouds are actually routers (hidden from the topology) allowing for integration of the different network topologies.
After I get all of the functions operating correctly on the server, I will post the topology file.
Monday, May 18, 2009
Daily Digest for May 19th
| Published Rough weekend |
| Published Daily Digest for May 18th |
| Working on documentation from home this morning. [#] |
| Heading out to Fremont for some security discussions with a customer. [#] |
| Back working on documentation. [#] |
Sunday, May 17, 2009
Rough weekend
We ended up not camping at all over the weekend. The park we were planning on staying at in Kansas City had received torential rain all day with another 2 - 3 inches predicted for Friday night. Instead of camping, we decided to go to World's of Fun on Saturday morning (leaving Omaha around 6:00AM) and then camping Saturday night only. While down there, we decided that it would probably be best to just head home after the park closes.
We decided this because it would be after midnight before we got the tents all setup and everyone in bed and settled down. We would then need to get up early Sunday morning, repack everything up and then head back home. As we got home a little after 1:00AM, that seemed to be a better plan.
We all had a great time at World's of Fun, but I did not get very far on the lab configuration. I will get the routers installed and configured tomorrow so I can start the testing process. I downloaded the VMware image for the AXP development platform. I just need to get a copy of VMWare workstation and a key from our rep and I will be ready to start the porting process.
We decided this because it would be after midnight before we got the tents all setup and everyone in bed and settled down. We would then need to get up early Sunday morning, repack everything up and then head back home. As we got home a little after 1:00AM, that seemed to be a better plan.
We all had a great time at World's of Fun, but I did not get very far on the lab configuration. I will get the routers installed and configured tomorrow so I can start the testing process. I downloaded the VMware image for the AXP development platform. I just need to get a copy of VMWare workstation and a key from our rep and I will be ready to start the porting process.
Saturday, May 16, 2009
Friday, May 15, 2009
Daily Digest for May 16th
| Published Working on Virtual Lab |
| Heading out to the gym. [#] |
| Working on network assessment documentation in the office this morning. [#] |
Thursday, May 14, 2009
Working on Virtual Lab
After messing around with the hardware configuration on my (previously working) Linux server, I finally got Ubuntu installed an booting on the system. Apparently, there is an issue with the configuration of the hard drives and the order of operation. I had a single 250GB IDE drive for the OS (which installed just fine) and three 500GB drives in an onboard SATA RAID configuration (one of the half-software and half-hardware RAID controllers). Going through the BIOS configuration (over and over and over and ...) I finally gave up trying to get the system to boot off the IDE drive first (no there was not an option for the boot order of the hard drives).
So, I ended up taking out one of the 500GB drives (I don't need the space on there anyway) and the IDE drive. Now, I have two 500GB SATA hard drives in a non-raid configuration. I used one for the OS, boot and swap. The other is primarily for storing the CD and IOS images, virtual machines and other related data for the lab configuration.
I will post a page on the lab server configuration soon. I am still working on the architecture, but the quad core CPU with 8 GB of RAM should be capable of handling everything I need for now. I plan on getting the rest of the configuration completed in the next few days (installing the ubuntu-desktop package now).
So, I ended up taking out one of the 500GB drives (I don't need the space on there anyway) and the IDE drive. Now, I have two 500GB SATA hard drives in a non-raid configuration. I used one for the OS, boot and swap. The other is primarily for storing the CD and IOS images, virtual machines and other related data for the lab configuration.
I will post a page on the lab server configuration soon. I am still working on the architecture, but the quad core CPU with 8 GB of RAM should be capable of handling everything I need for now. I plan on getting the rest of the configuration completed in the next few days (installing the ubuntu-desktop package now).
Daily Digest for May 14th
| At the gym. [#] |
| Published Daily Digest for May 14th |
| Heading baxk to Iowa for the morning. [#] |
| Heading back to omaha [#] |
| Taking Sami to dance class. [#] |
Wednesday, May 13, 2009
Daily Digest for May 14th
| Published Daily Digest for May 13th |
| Heading to iowa to finish the network assessment data collection. [#] |
| Published Cisco Discovery Tool |
| Heading home for the day. [#] |
Cisco Discovery Tool
Today, I am in Iowa working on a network assessment for a hospital. As part of the network assessment, I am running the Cisco Discover Tool which is supposed to detect Cisco and some non-Cisco devices on the network via SNMP and ping sweeps across the network.
So far, the process has been pretty seamless, but I had to run the discovery twice now. I think this is because I am using the discovery tool in the Microsoft Virtual PC virtualization software as opposed to running the application native. When I tried to run the application on my native Vista x64 system, I got some messages about the software not being supported on my operating system. My virtual PC is just a standard Windows XP with SP3 installed (and minimal other software enabled).
I was able to get the Virtual PC to run across my SSL VPN to my home network. It was able to discover my one Cisco ASA 5505 firewall and had no issues. I will try it again when I get my new lab server up and running tonight (with dynamips).
So far, the process has been pretty seamless, but I had to run the discovery twice now. I think this is because I am using the discovery tool in the Microsoft Virtual PC virtualization software as opposed to running the application native. When I tried to run the application on my native Vista x64 system, I got some messages about the software not being supported on my operating system. My virtual PC is just a standard Windows XP with SP3 installed (and minimal other software enabled).
I was able to get the Virtual PC to run across my SSL VPN to my home network. It was able to discover my one Cisco ASA 5505 firewall and had no issues. I will try it again when I get my new lab server up and running tonight (with dynamips).
Tuesday, May 12, 2009
Daily Digest for May 12th
| Finished up early, making dinner now (no lunch, so I am starving) [#] |
| Heading to the gym. [#] |
| Published Just got blog up! |
| Grabbing lunch from Runza with Michelle. [#] |
Just got blog up!
I just got the blog up and running. I am still working on the rest of the site. I will post curent information and status of the items I am working on here.
Subscribe to:
Posts (Atom)