|
Currently we have a larger mail system but only 500
mailboxes have tmda installed. We are using qmail. Python (specifically TMDA) seems to thrash the CPU when it
is called. This is not specifically horrible but is a concern if the
number of mailboxes reaches thousands or more. I am currently using plain text lists and am wondering if
using CDB would clear up the problem or at least make it considerably better. Is there currently any good documentation on going from a
plain text type of situation to using CDB for the list files? I am also using tmda-cgi which installes TMDA on a mailbox so any info on how to get tmda-cgi to install using CDB instead of plain text files. Also, is it practical to think this would really help much? Could there be another way to minimize the “thrashing”
of the cpu? We currently use qmailscanner
also, written in perl, which scans every single email
and the processing time and load overall is far less than what python is taking
(tmda). Considering that the qmailscanner
is doing hundreds per second when TMDA is doing a couple per second tmda is something to look at. Sam |
_____________________________________________ tmda-users mailing list ([EMAIL PROTECTED]) http://tmda.net/lists/listinfo/tmda-users
