I have a client that uses a cloud application and the users are complaining about sporadic performance problems during certain operations within the program. Using all the bandwidth and network traffic monitoring that we have in place there are no obvious issues with the Internet connection. In fact, it appears they aren’t using anywhere close to all the bandwidth, especially in this application. The cloud provider has a speed test that can be run to test the connection between us and their data center. When the users are experiencing the performance issues, I have them run the speedtest. The results the users are getting back are showing a very healthy connection; always 25ms or less of latency and greater than 50Mbps upload and download. In addition to this, I have made packet captures of the cloud application (during times where no performance issues are noticed) and I have seen that these operations involve very little data transfer. In most cases it is less than 3MB in total data for an operation that when the performance problem is present is taking as long as 5 minutes to complete. In many cases the users are even having to “end task” on the cloud application because, from what they tell me, it looks as if the application is locked up.
Currently we seem to be stuck in finger pointing mode in that the cloud provider (and some users) are just blaming our “bandwidth” as the problem. I feel like the results that the users are getting from the speedtests they are running while the problem is occurring are conclusive proof that there isn’t an issue with bandwidth/connectivity but I am still struggling to get traction with the cloud provider.
Does anyone else have any suggestions on what I can do to isolate the problem and/or prove make quality progress on getting the issue worked on in a quality way by the cloud provider?
Any suggestions are greatly appreciated and please let me know if you need more info from me on this.
No comments:
Post a Comment