Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-17 Thread Denys Smolainiuk
Great! I've added myself to this ticket as a watcher. Thanks, Denys Smolianiuk Guten Tag Denys Smolainiuk, am Donnerstag, 16. August 2018 um 17:31 schrieben Sie: > Any thoughts? Thanks for sharing your research, I've additionally created a ticket to track your patches: https://issues.apache.

Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-17 Thread Denys Smolainiuk
Hi Matt, Don't think boost has queue following disruptor pattern. But hopefully we don't need multiple consumers/producers here. Otherwise I'm glad to here I'm not the only one seeing such issues :) Thanks, Denys Smolianiuk I don't know much about the state of log4cxx architecture, but nearl

Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-17 Thread Thorsten Schöning
Guten Tag Denys Smolainiuk, am Donnerstag, 16. August 2018 um 17:31 schrieben Sie: > Any thoughts? Thanks for sharing your research, I've additionally created a ticket to track your patches: https://issues.apache.org/jira/browse/LOGCXX-500 Mit freundlichen Grüßen, Thorsten Schöning -- Thorst

Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-16 Thread Remko Popma
All these sound like very nice improvements! On Fri, Aug 17, 2018 at 4:13 AM, Robert Middleton wrote: > Interesting results! I haven't experienced any problems with logging > myself, but I've also never gone into benchmarking it. > > I don't have much experience with memory pools, but if you're

Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-16 Thread Robert Middleton
Interesting results! I haven't experienced any problems with logging myself, but I've also never gone into benchmarking it. I don't have much experience with memory pools, but if you're still getting blockages on a call to new/malloc you may be able to get around it by using the memory pool and t

Re: [log4cxx] Logging in Timing-Critical Applications

2018-08-16 Thread Matt Sicker
I don't know much about the state of log4cxx architecture, but nearly all your points (other than the lock in stringstream) are points we optimize for in log4j2 at least. Even the stringstream optimization sounds similar to StringBuffer versus StringBuilder in java. As for the queue used in async l

[log4cxx] Logging in Timing-Critical Applications

2018-08-16 Thread Denys Smolainiuk
Hello All, I'd like to share some experience as well as some patches with regard to using log4cxx in timing-critical application. First few words about our requirements: it's a service which must generate some network packets with up to hundred of microseconds precision. Thus, it's very impor