Alexander Gallego
Alexander Gallego
https://queue.acm.org/detail.cfm?id=2209336 we have an implicit fifo queue w/ the future impl. We need to timestamp the requests at time of accept() w/ the low precision timer (10ms steps) and then...
I just added the american fuzzy lop to fuzz the inputs to the wal readers. Need to fix the readers and writers of this stuff.
somethign like ``` option (rpc_core.method_default_deadline_ms) = 5000; ``` annotation on the method. we already do this w/ defaults of i think some seconds but we can code gen it -...
In order to be exception safe: https://github.com/smfrpc/smf/blob/master/src/core/rpc_server.cc#L169 use seastar::semaphore_units()
currently we `flush()` to the socket with ever RPC. This makes the RPC low latency but we give up all the throughput. We could have a timeout style check. if...
in core memory pressure needs a more holistic view. that's not a good use of the limits semaphore
``` struct connection_callbacks { void connection_open(const ...&); void connection_error(const ...&); void connection_exception(const ...&); void will_parse_header(); void did_parse_header(const &).... void will_parse_body() void did_parse_body(const &...) void connection_set_error()... void connection_disable.... void connection flushing...
https://github.com/facebook/treadmill/blob/master/Scheduler.h
unfort toplev uses python 2.7 and needs tk bindings https://github.com/andikleen/pmu-tools/wiki/toplev-manual apt-get install python-tk-dev ? for fedora is `python3-tkinter.x86_64 ` yield a new readme w/ links to: https://software.intel.com/en-us/vtune-amplifier-help-tuning-applications-using-a-top-down-microarchitecture-analysis-method add perf record...
notifications: email: - [email protected] - [email protected]