Any upstream size data?

[danlisarice]

Do you have transfer size for the upstream direction for all the requests? I assume the transfer size charts are just the data going from the server to the browser?

[rviscomi]

Hi Dan. Yes the transfer size charts like this are referring to the average download size:

If you wanted to find out the size of the requests themselves, you could query the requests table like this:

SELECT AVG(reqHeadersSize) FROM [httparchive:runs.2017_05_15_requests] WHERE mimeType = 'application/javascript'

The result is 448.7 bytes.

[pmeenan]

Not sure if it is in the SQL tables but the HAR files have a bytesOut measurement (though it is usually a tiny rounding error in the grand scheme of things).

[andydavies]

Have seen clients where upstream is an issue - mainly due to the size of cookies attached to each request

…

On 9 June 2017 at 17:19, Patrick Meenan ***@***.***> wrote: Not sure if it is in the SQL tables but the HAR files have a bytesOut measurement (though it is usually a tiny rounding error in the grand scheme of things). — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#114 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAaELE-MQvDW4fEoHVMHOk58KMZOHwCYks5sCXCRgaJpZM4N1eCm> .

[rviscomi]

Yeah, here's a relevant thread from the forum: https://discuss.httparchive.org/t/size-of-document-cookies/49. Also https://discuss.httparchive.org/t/1mb-of-http-overhead-due-to-towerdata-cookies/479

[danlisarice]

Thanks I will do some queries and poke around in the data. Access networks can be pretty asymmetrical with different US and DS loss attributes and I am noodling on upstream packet sizes, counts and the impact on physical layer efficiencies for a variety of different traffic loads including web surfing.

[rviscomi]

Looks like this is resolved. Closing.