3 ms·
If you run into contention issues have you tried scaling horizontally? Running on 128 single core VMs should mitigate what you see.
by tjungblut 5y ago
If you run into contention issues have you tried scaling horizontally? Running on 128 single core VMs should mitigate what you see.
- CyberDildonics 5y agoThat would turn communicating between threads into communicating between VMs, which is a disastrous change in itself.
- tjungblut 5y agohow do you know the threads need to communicate with each other?
- CyberDildonics 5y agoHow do you know they don't?
- gpderetta 5y agoIt would. On the other hand multiprocessing with explicitly shared memory could be a fast (but potentially painful) workaround.
- CyberDildonics 5y agoThat could make sense for processes, which would also get around the exception unwind lock, but I don't know how that would work with every thread moved into a separate VM.
- gpderetta 5y agoOf course. Using VMs here would very deep into overengineering territory.
- nyanpasu64 5y agoIf you want to avoid a global process lock, multiple processes will have less orchestration and communication overhead than multiple VMs (though still a lot more than necessary with threads within a process).