Hi! On Tue, 17 Sep 2013 15:59:28 +0200 Joachim Wiedorn <[email protected]> wrote: > For me it isn't clear why this problem (still) exist. I will try to find > the reason.
We just had the same problem. In /var/tmp were lot's of files (~15000; 77GB). I digged into the code of squidguard and libdb and for me it looks like if squidguard is using an unnamed in-memory btree via libdb. If the amount of data still fits into the in memory cache, no file is created. If too much data is stored within the btree, a temporary backing file is created in /var/tmp (if nothing else is defined in DB_ENV). Currently I have no idea who is responsible for the backing file. If I'm right, squidguard does not get an file handle for the backing file and is therefore unable to take care of it. It is possible to define a different tmpdir for the backing files but e.g. /tmp is no real solution since we have multiple squidguard instances running which are quite often restarted. On each start of an instance, a new backing file is created. Any ideas on how to handle it? Did I miss something? Regards Sönke -- Sönke Schwardt-Krummrich Open Source Software Engineer Univention GmbH be open. Mary-Somerville-Str.1 28359 Bremen Tel. : +49 421 22232-41 Fax : +49 421 22232-99 [email protected] http://www.univention.de Geschäftsführer: Peter H. Ganten HRB 20755 Amtsgericht Bremen Steuer-Nr.: 71-597-02876
signature.asc
Description: This is a digitally signed message part.
