Re: libosmocore stderr log performance / blocking

Dear list, Vadim requested to also compare syslog based logging. In theory, I would have expected it to perform similar to gsmtap, given that it also doesn't do anything else but sending UDP packets. However, the performance looks slightly worse than with gsmtap: @usecs: [2, 4) 1167 | | [4, 8) 2056285 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@| [8, 16) 8403 | | [16, 32) 1880 | | [32, 64) 6 | | [64, 128) 6 | | [128, 256) 0 | | [256, 512) 2 | | [512, 1K) 1 | | [1K, 2K) 738 | | [2K, 4K) 2 | | [4K, 8K) 0 | | [8K, 16K) 0 | | [16K, 32K) 0 | | [32K, 64K) 0 | | [64K, 128K) 0 | | [128K, 256K) 0 | | [256K, 512K) 0 | | [512K, 1M) 0 | | [1M, 2M) 0 | | [2M, 4M) 0 | | [4M, 8M) 0 | | [8M, 16M) 0 | | [16M, 32M) 0 | | [32M, 64M) 0 | | [64M, 128M) 0 | | [128M, 256M) 0 | | [256M, 512M) 0 | | [512M, 1G) 0 | | [1G, 2G) 49 | | Note the significant count of samples in the 1..2ms bucket. That's still quite a lot, compared to the usual 4..8us. Reminder: With gsmtap loggingwe had no log lines with more than 128us delay. The results are reproducible, I just double-checked that gsmtap defintely outperforms syslog in terms of the maximum delay/latency caused by logging. -- - Harald Welte <laforge(a)osmocom.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)