On Tuesday 15 October 2019 15:01:14 Dr. Nikolaus Klepp wrote:
Hi Gene!
Are you sure your hardware is OK?
Nik
drives are fairly fresh as I always install to a new drive so I can get stuff off the old one, cmos battery replaced just a few months ago. Nothing is complaining. 29days uptime. kmail is using much of a cpu but not 100% continuous as I've no doubt it busy rebuilding indice files yet. crash at restart after sorting stuff for about a minute. crashed, reopened instantly and now I expect it will idle back when the indice recreation is finally done. the gkrellm display now looks like a normal but busy machine, no big green 100% blocks jumping from core to core.
Anno domini 2019 Tue, 15 Oct 13:00:44 -0400
Gene Heskett scripsit:
Greetings all;
Are you send those to /dev/null? I've sent a whole bunch of them.
These crashes always seem to be preceded by a couple days of a kmail session burning up a core of 4 cores, bouncing to the next available core, at 99 to 100% at 15 second or so intervals.
These seem to be related to kmail finding trash files in its database, causing problems while re-indexing. I have two folders which are subfolders with a given years messages manually sorted into that years corpus.
When I relaunch kmail after one of these crashes, I am getting advisories that so-and-so has an index problem, and its generally the top level folder, and 2 or 3 of its subfolders that are named. And the older folders are slowly being emptied, as in messages are disappearing despite having no expiry set up.
Can anything be done, or is there something I can check-uncheck someplace?
Thanks.
Cheers, Gene Heskett
Cheers, Gene Heskett