httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Tao <>
Subject 0.73g + FreeBSD 2.0.5
Date Fri, 23 Jun 1995 06:40:53 GMT
On Thu, 22 Jun 1995, Chuck Murcko wrote:
> Built 0.7.3g on Solaris 2.3, started it with 15 children. Ran about 10
> minutes, then got a raft of SIGSEGVs and kernel panic with a data fault.

    I had the same problem (bad page fault, kernel panic) with 0.73g
the first time I ran it on my FreeBSD machine, but not on the second
try.  It looks like a FreeBSD issue, so I won't bother this list with
the details.  20 children, 50 clients, died in less than 20 minutes.

> [Thu Jun 22 14:17:27 1995] socket error: accept failed
> [Thu Jun 22 14:17:27 1995] could not get local address
> [Thu Jun 22 14:17:27 1995] httpd: caught SIGSEGV, dumping core

    I've never seen any of the messages in my logs, despite crashes
due to various other reasons.  The pattern goes something like this:
within the first few minutes, two to three children are zombified.
Then one or two more go in the next ten minutes or so, and after that
more pop up at a more or less even pace.  Note that this is with 10+
requests per second, so under less strenuous conditions, the zombies
may take much longer to show up.

    Last night, I ran the same test as before (20 children, 50
benchmark clients) but with a cron job sighupping the parent server
once an hour.  From the client logs, it looks like a SIGHUP kills any
connections currently in progress.  Is this still true?

    Came in this morning and the machine was still chugging away, no
fd leaks, CGI's were happily running, not too much swapping going on,
but had three zombies (this was 8 minutes after the last SIGHUP).

> And now, on a lighter note:
> Documentation is like sex: when it is good, it is very, very good; and
> when it is bad, it is better than nothing.
> 		-- Dick Brandon

Brian ("Though this be madness, yet there is method in't") Tao <-- work ........ play -->

View raw message