-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 13/02/2006 14:01, Ivan A. Beveridge wrote:
>>> plugin_pipe_size: 10240000 >>> plugin_buffer_size: 10240 > > OK - will try this. I'm not sure what pps I'm getting, but we're > sampling ~30Gbps traffic currently (that'll rise to 90Gbps soon-ish and > certainly upwards in the coming months). That would be across 2 > (probably) memory plugins. > I'm going to try adding the above suggested plugin settings for the > moment, and will probably try compiling & using the rc2 snapshot you put > out. It looks like I need to tweak something else for the backtrace to > be of use though - let me know what / thoughts! A little update. The problem _may_ have been related to the number of 'pmacct' processes hanging around (1 running every 5mins). This is related to the number of RRD files that have to be updated at each pass .... currently over 16000, which is proving a bit of a struggle on the test setup, time-wise. Looks like we're going to need to look at some optimisations for our processing (not pmacct stuff). Cheers Ivan - -- Ivan Beveridge <[EMAIL PROTECTED]> http://www.linx.net/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFD9IH6QQZN5jq7vncRAqZjAKCgKOC6ARNUAj8yVcgWXxq1SxBjUQCdEF/x wZyuVyuasvu9p6iOAWHXraw= =vfjc -----END PGP SIGNATURE-----
