OK, the person came back, I got him to disable sendfile, and apache
started transferring at full speed.
I believe sendfile should be disabled by default at least on Windows.
Rici
On 17-Mar-05, at 11:48 AM, Rici Lake wrote:
Bill,
What's the performance problem?
There was someone on #apache yesterday who claimed to experience the
following problem: On a local switched network, Apache2 on Win2K3
transferred at 800kbps, while IIS on the same box transferred at
8-11mbps. When he replaced the cable to the server with a shorter one,
Apache speeded up to 4 mpbs but IIS still ran at full speed. He tried
it with both Windows and Mac OS X clients, and experienced the same
symptom.
Could this be related to sendfile?
Rici.
On 17-Mar-05, at 11:12 AM, Bill Stoddard wrote:
Joshua Slive wrote:
Jeff Trawick wrote:
IMHO, all default conf files except for highperformance[-std].conf
should have the "EnableSendfile off" directive uncommented, under
the
theory that people who want the extra performance can take an action
to re-enable sendfile and then pay attention to the web server
behavior afterwards but people who just want the darn thing to work
shouldn't be bothered. Some percentage of the latter group will
have
problems they are ill-equipped to handle, because some feature that
is
not important to them is enabled.
We've discussed this before and nothing ever happened. I'm willing
to
bet that the experiences of a few more unlucky users since the last
discussion may have shifted the balance. Comments?
+1, as before.
From the users' perspective, sendfile results in unexplained
corruption or uninterpretable error messages pretty much any time a
network filesystem is used to host content (and random other times
on win32).
Hey Joshua,
Guess I've not been paying attention... what type of problems have
you seen with sendfile (TransmitFile) on Win32? We do have a
performance problem (will leave out details unless your interested)
but I was not aware of corruption problems...
Bill
