jlewis@lewis.org wrote:
>On Sat, 26 Jun 1999, Stephen R. van den Berg wrote:
>> That means, total leaked is approximately: 142550-46504-10000 = 86046 KB
>Included below is the full general info from that previous post. What's
>the 10000 above? His 7mb of free blocks plus ~3mb for the squid
No. The 10MB is my very rough estimate on his real malloced data not part
of the leak.
>If I take 226629 - 96695 - 12000, I get 117934KB. If I ignore ICP,
>117934KB * 1024 bytes/KB / 3197449 http reqs = 37.8 bytes/req
> Number of HTTP requests received: 3197449
> HTTP requests per minute: 1528.1
> Ordinary blocks: 217882 KB 353099 blks
> Total accounted: 96695 KB
> Number of HTTP requests received: 895997
> HTTP requests per minute: 655.6
> Ordinary blocks: 35915 KB 14871 blks
> Total accounted: 26731 KB
Comparing these, indeed, I can only conclude that the second one does *not*
have the leak. I.e. the worst (or most optimistic) leak is:
9184/895997*1024 = 10.5 bytes/request
Not even close to 36.
Are you certain the smaller server experiences the leak as well eventually?
--
Sincerely, srb@cuci.nl
Stephen R. van den Berg (AKA BuGless).
"Father's Day: Nine months before Mother's Day."
Received on Sat Jun 26 1999 - 14:35:51 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:47:00 MST