The design of Analog is such that it shouldn't use noticably more memory 
than the executable size plus the size of the log file(s) and/or cache 
file(s) (and usually far far less). There is only one exception that I 
can think of, the dnscache file is loaded into RAM on startup, whatever 
it's size.

Jason


On 8/17/02 12:31 PM Kai Schaetzl ([EMAIL PROTECTED]) wrote:

>We have one machine were memory is scarce (128 MB) but analog takes 
>about 50 MB of that (sometimes even more) each time it runs. This is 
>causing problems. We have a fairly heavy-traffic site on that 
>machine, so I skimmed the config for that site down, especially by 
>switching the HOST report off, thinking this would solve the problem. 
>However, the whole analog phase for the few sites on that machine 
>still took longer than I think it should and still caused problems 
>for other processes, although the run for that site is now fairly 
>short. So I looked deeper at each of the runs and found that analog 
>uses those 50 MB of memory EACH time it runs, no matter how big the 
>log file, even for tiny ones of 2 KB, even if just doing a 
>summary/file type/file size report (traffic overview) from cache 
>files. I don't quite understand why it is taking so much space for so 
>small files or still needs that much memory for extracting data from 
>the cache files. I don't remember this behavior from older versions. 
>This isn't normal isn't it?
>Also, I remember from old 3.x and 4.x days that there were explicit 
>options to limit the memory analog uses instead of those LOWMEM 
>things which don't *really* limit in a foreseeable way. Isn't there 
>really no way to say "take 25 MB and no more, never"?
>Just upgraded from 5.0 to 5.24 but it still grabs that many memory. 
>This is a Red Hat Linux 6.1 system with 2.2.19 and fairly new libs.


-----------------
[EMAIL PROTECTED]
-----------------
Dr. Seuss books . . . can be read and enjoyed on several levels. For
example, 'One Fish Two Fish, Red Fish Blue Fish' can be deconstructed
as a searing indictment of the narrow-minded binary counting system.
  -- Peter van der Linden, Expert C Programming, Deep C Secrets


+------------------------------------------------------------------------
|  This is the analog-help mailing list. To unsubscribe from this
|  mailing list, go to
|    http://lists.isite.net/listgate/analog-help/unsubscribe.html
|
|  List archives are available at
|    http://www.mail-archive.com/[email protected]/
|    http://lists.isite.net/listgate/analog-help/archives/
|    http://www.tallylist.com/archives/index.cfm/mlist.7
+------------------------------------------------------------------------

Reply via email to