K6 TCP segment size, is it configurable?

Hi @fimal, welcome to the community forum! :tada:

You likely can not or at least I don’t know how that could be done.

From the looks of it you are hitting the issue here Strange k6 behaviour with HTTP2 and the thing is not that it conforms to MSS or something like that, but that the http2 implementation in golang will at most “buffer” 4kb before sending them.

Later the kernel/tcp stack might segment it further but in this case this isn’t needed.

I haven’t communicated this with the go team, sorry it has fallen through the cracks at some point. It also seems like either the constant will get bumped or it will be bogged down in a lot of discussions.

Hope this helps you understand it, but I have no workaround for you, sorry