As part of my move from 1.6.22 on Solaris 7 to 2.2.4 on Solaris 10, I need to bulk move about six hundred mailboxes totalling about 120GBytes. For the pilot users I've used imapsync, which works correctly (including moving flags) but is quite slow and resource intensive.
For service use, I'd ideally just rsync the data over and reconstruct the mailboxes on the receiving side. On current showing, this would be a _lot_ faster. However, it appears to lose flag settings other than seen, which some users will grumble about. I think this is because cyrus.seen is flat, while cyrus.index is a database --- I presume BerkeleyDB, but I'm not entirely sure. I compiled 1.6.22 using BerkeleyDB 3.1. I'm using Berkeley 4.1 and Skiplist in 2.2.4. If I compiled a copy of cvt_cyrusdb using the same set of BerkeleyDB headers and libraries, would it then be capable of converting cyrus.index? Or would the differences between 1.6.22 and 2.2.4 be too great? It worries me that a copy of cvt_cyrusdb which understands Berkeley 4.1 won't look at the files, because I was under the impression that there was back-compatibility there. ian --- Cyrus Home Page: http://asg.web.cmu.edu/cyrus Cyrus Wiki/FAQ: http://cyruswiki.andrew.cmu.edu List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html