Hello!

I run pmacct on FreeBSD 9.1/9.2 with mysql. Sometimes on on of my
systems the following happens:

#tail -f /var/log/pmacctd.log
Aug 14 14:22:42 WARN: Failed during write: No buffer space available
Aug 14 14:22:42 WARN: Failed during write: No buffer space available
Aug 14 14:22:42 WARN: Failed during write: No buffer space available
Aug 14 14:22:42 WARN: Failed during write: No buffer space available
Aug 14 14:22:42 WARN: Failed during write: No buffer space available

During that pmacctd is using almost 100% of one of CPU cores, and data
is not stored to mysql database.

Restarting pmacctd helps, for a random period (several days).

My pmacctd.conf:
daemonize: true
interface: lan
pidfile: /var/run/pmacctd.pid
interface_wait: true
logfile: /var/log/pmacctd.log
post_tag: 1
aggregate[in]: dst_host
aggregate[out]: src_host
aggregate_filter[in]: dst net 10.0.0.0/8
aggregate_filter[out]: src net 10.0.0.0/8
plugins: mysql[in], mysql[out]
plugin_pipe_size: 20971520
plugin_buffer_size: 10240
sql_host: 176.111.128.5
sql_db: pmacct
sql_table: acct_rx_opole
sql_table_version: 1
sql_passwd: mdx2WxmpLcR3erY5
sql_user: pmacct
sql_refresh_time: 600
sql_dont_try_update: true
sql_optimize_clauses: true
sql_history: 10m
sql_history_roundoff: mh
sql_multi_values: 1024000
sql_max_writers: 15
sql_preprocess: qnum=50, minp=25

Any help would be appreciated :)

Regards,
Krzysztof



_______________________________________________
pmacct-discussion mailing list
http://www.pmacct.net/#mailinglists

Reply via email to