As I have recently found out. Avid PCoIP fans in most cases will ‘always’ disput any theory on packet loss or latency with their protocal by calling it a ‘myth’ (via their blog). Unfortunately we all just cant take their word for granted especially when deciding whether or not to invest in the technology.
One other question to ask is why is there no publically availible Software Development Kit [SDK] or documented API? As far as I’m concerned, its completely unfortunate but there there are some things we can do to watch the packets drop and help us determine what is an acceptable range for latency when dealing with the PCoIP protocal.
Since we now have the above past us, lets write some .NET code and allow our automation to tell us when a packet drops or latency is above our predetermined threshold. Its quite possible you could use a vbscript, powershell or cmd file to perform the same analysis, but a managed environment is more suatible for us in this scenario. Now lets establish our requirements and delve directly into the code.
A standard ping will not suffice for this this test, we want to use something a little more robust and capible of determining error specifics via our custom logic. By using system.net.sockets, we can deeply analyse error codes such as port issues, hardware errors, bad routes/requests, detect mtu changes and other general failures. In the event you dont already have a list of host or portal IP’s we want to have the ability to drop in a subnet and find all applicable PCoIP devices.
Next we want to set our thrshold to alert to do a datagridview when the condition is met. We also want to send our alert to another datagridview when we detect packet loss under our predefined conditions. Then we simply want to just add a timer and set it to how often we want to perform the analysis on the hosts and portals.
Because we like to share the love – this software is completely free [download link is listed below]. We want our readers to be informed and educated decision makers. After you test the software and perhaps would like to see any additional features added, please send us a note and should it be a value added feature – we will add it and give you credit. Please enjoy and let us knowwhat you think!
- Here is what our code looks, fully functional and ready to perform analysis.
- To recursively and quickly search a specific subnet for pcoip compatable hosts – key in your subnet, starting address and ending address. And if you want (optional) set your criteria or threshold for analysis against packet latency. Its important to know what to expect in terms of latency and where you are performing the analysis. Unless you have a completely amazing link between your deskt in Canada and the portal in London you will have to adjust accordingly. Then use the data provided in this application to establish a baseline.
- When you are ready to begin the analysis – click on the blue “begin analysis” button
- Once the packet analysis begins, any detected errors, drops or reported issues will be reflected in the “analysis results” tab. The timer will peform the report against each entity within the “subnet analysis” tab.
- Now when you see packets which have a latency higher than your predetermined level, they will be reported in the “threshold” analysis tab.
