On 4/7/2016 11:53 AM, James Peach wrote:
On Apr 7, 2016, at 9:33 AM, Susan Hinrichs <shinr...@network-geographics.com>
wrote:
I'm testing congestion_control to enforce per-origin connection limits. In my
micro tests, all is looking good. However, as I'm reviewing jira for open
issues, I came across the following comment from Leif in TS-1221
"I'm not sure, but I think congestion control code is missing vital parts that was
never open sourced (such as the ARM kernel module). But again, I'm not 100% sure. Does
anyone know? If it's crippled, we should make sure to clean up the configs and
traffic_line etc."
Does anyone have definitive information the state of congestion_control? Is
anyone using it these days?
AFAIK no-one has ever used it, but Faisal was asking about a per-orgin
bandwidth control over in users@, so I suspect the capability would be useful
to people. Do you think it could be salvaged?
J
I've played with the max_connection settings for congestion_control so
far. That mechanism returns an error right away rather than delaying
and spinning as the origin_max_connection feature does. Looking through
the code, things seem reasonable so far. I'll push on with caution.
I think what Faisal is asking for is better served by using TC. I
replied on that thread.
