Open williammck opened 10 months ago
What's the max size of payload that you're sending here? It may be bandwidth exhaustion from very large payloads, but that is quite the timeout happening - any idea what size you're sending?
Depends on the key - each value is up to a kilobyte max, but usually hovers around 64B - 128B. We are sending up to a few thousand of these all at once. But even pipelined, and with some headroom, I'd be shocked if we're upwards of a megabyte in one go. What's weird too is it's been behaving a lot more like normal this week, but we're seeing the same amount of load, if not more, on the service and the node it's living on. I'm a bit puzzled as to why that'd be happening.
Hey folks,
I've been battling recent issues with one of our workers regularly getting Redis timeouts and would appreciate some help.
A bit of overview:
FsdService.AddToRedis
, which runsStringSetAsync
a couple thousand times.for
loop over each client callsStringSetAsync
twice, so in that example, 1500 calls.FsdService.AddToRedis
reads from.I've tried upping the threads in the pool up a bit (set
ThreadPoolMinThreads
to 16), but that didn't seem to help much. Should I increase it further?I've looked into SuperSimpleTcp and it seems to be doing everything in tasks/async, so I'd hope we're not blocking, but I can't quite tell.
Any ideas or things I should look at?
Stacktrace examples: