Michael Mayer wrote: > [EMAIL PROTECTED] wrote: >> But if you can empirically demonstrate that over-indexing >> signifficantly slows down the typical (mostly read) workload of >> DBMail, I'd be most interested to see it. > > I just decided to test it... maybe on the weekend :) > > The only question is how to set up the test environment, because my own > dbmail database is too small. Can somebody give me a hint how to produce > test data or even send a dump?
If you have a large set of email from various sources (spam/ham/plain-text/attachments) you should insert those into dbmail using something like: #!/bin/sh i=1; j=1; while [ $i -lt 100 ]; do while [ $j -lt 100 ]; do cat somebigmailbox.mbox|\ formail -ds dbmail-smtp -u testuser$i -m mailbox$j let j=$j+1 done let i=$i+1 done that should produce a nice set of mailboxes spread acros multiple accounts. > What are the 5 most common queries for an average dbmail server? Difficult. I don't have any numbers. Mostly depends on the most prevalent clients. You can find some of the most used IMAP commands for TB and Outlook in test-scripts/textimap.py in the testFetch testcase. -- ________________________________________________________________ Paul Stevens paul at nfg.nl NET FACILITIES GROUP GPG/PGP: 1024D/11F8CD31 The Netherlands________________________________http://www.nfg.nl _______________________________________________ DBmail mailing list [email protected] https://mailman.fastxs.nl/mailman/listinfo/dbmail
