> On Tue, 6 Aug 2024 18:12:25 -0400 > ma...@mohawksoft.com wrote: > >> Without getting into too much detail, their SVC server product had a >> real-time polling process that maintained timers on various >> processes, if the processing took too long, the system would >> "fail-over" to the other node. They were getting too many timeouts >> with hyper-threads enabled. > > Interesting. I suppose this goes hand in hand with dynamic frequency > scaling (CPU Turbo Boost) causing high thermal loads and power > throttling under sustained compute loads.
Yup, that was one of the issues. _______________________________________________ Discuss mailing list Discuss@driftwood.blu.org https://driftwood.blu.org/mailman/listinfo/discuss