HTTP Throughput inconsistent

[blkwomenread]

Hello. My org sees inconsistent HTTP or Web throughput for our tests in ThousandEyes. Do you know why this would be? Some of the tests show throughput for some time and then it stops completely or just nothing at all. I compared a few tests and they're configured exactly the same. Perhaps it's just the URL that it's hitting that may be causing this.

[Tyler Langston · Accepted solution]

Hi @blkwomenread - thanks for your question! I ran this by our experts and they had some insight:

We would need some more information to understand why the throughput is inconsistent.

We calculate Throughput by measuring the amount of data that was received from the target HTTP server, and dividing that by the time it took to receive all of the data. 

To better understand the cause of the variance we would need to look at the network holistically, not just the web layer. Perhaps comparing the agent to server metrics like end-to-end loss, latency, jitter. And at the same time, comparing any changes in the path visualization between test rounds as sometimes latency can be introduced by excess hops in the network path between the source and target.

To give specific guidance on their situation we recommend they reach out to our support channels so we can take a look at the test settings as well to see if anything can be tweaked.

Based on their input I recommend you contact our support engineers and work with them on finding a root cause. The best way to reach them is via chat - available 24/7, with an average speed of answer of 15 seconds! Specifics on that process here. 

Thanks!