Re: [ic] Re: ALERT: bad pipe signal received for /page.html

[Josh Lavin <suppressed>]

On Dec 11, 2006, at 5:35 PM, Ron Phipps wrote:
> Josh Lavin wrote:
> > On Dec 11, 2006, at 12:42 PM, Ron Phipps wrote:
> > > Josh Lavin wrote:
> > > > On Dec 11, 2006, at 12:02 PM, Grant wrote:
> > > > > > Hello, I've been plagued by apache2 segfaults ever since I  
> > > > > started
> > > > > > using Interchange::Link years ago.  The latest Link.pm has ALERT
> > > > > > messages accompanying the segfaults in error_log:
> > > > > >
> > > > > > ALERT: bad pipe signal received for /page.html
> > > > > > [Sat Dec 09 10:27:55 2006] [notice] child pid 21337 exit signal
> > > > > > Segmentation fault (11)
> > > > > >
> > > > > > Does anyone have any advice on solving this?  I'm using  
> > > > > apache-2.0.58
> > > > > > and mod_perl-2.0.2 in Gentoo Linux.
> > > > >
> > > > > The segfaults are eliminated by commenting out the $r stuff in the
> > > > > die_page sub.  I still get the ALERTs though.  Does anyone have  
> > > > > any
> > > > > advice on figuring out why I'm having the bad pipe problem?  Is  
> > > > > there
> > > > > an easy way to add extra debugging info to the sub?
> > > > >
> > > > > Also, restarting IC with PERL_SIGNALS=unsafe increases the  
> > > > > ALERTs 50 fold.
> > > > I've been seeing this too, on my Apache 2 and latest Link.pm. I  
> > > > also had to use PERL_SIGNALS=unsafe and so I get quite a lot of  
> > > > these.
> > > > The visible effect on the browser is that the page or image  
> > > > (which Link.pm apparently still has some part in delivering)  
> > > > does not load. I get them myself when browsing and testing my  
> > > > websites, and I have never stopped loading a page or had any  
> > > > other problems on non-IC sites I host.
> > > > I was told the problem stems from either the browser and a stop  
> > > > button or some other network fault. I may go back to Apache 1.3  
> > > > to get around this.
> > >
> > > I saw this occur on two different installations about 4 months  
> > > ago.  It was suggested that I abandon the use of Link.pm and go  
> > > back to using the cgi method with URL rewrite rules as this was  
> > > just as fast and proved stable over the years.
> > Is plain CGI really as fast as mod_perl or mod_interchange? That'd  
> > be my only concern with switching back to CGI+rewrites.
> > Maybe it is ok -- see mod_interchange's README:
> > "The Interchange link protocol has been
> > implemented via an Apache module which
> > saves us the (small) overhead
> > of the execution of a CGI program."
>
> I have not done any a/b tests to determine which is faster, others  
> have.
>
> Vlink/tlink is currently being used without any speed issues during  
> high traffic periods at:
>
> www.bcstore.com
> www.frozencpu.com
> www.citypass.com
> www.reliablemedical.com
>
> Quite a bit of time was put in with BCStore to compare which method  
> was stable and fast and it was determined that tlink/vlink was the  
> way to go.
>
> I thought that we needed to use mod_interchange/mod_perl for  
> FrozenCPU, but quickly found out that under Apache 2.0/mod_perl the  
> issues far outweighed any perceived speed up.  tlink/vlink are  
> compiled c programs and should be very quick in execution.
>
> At End Point we use rewrites with tlink/vlink pretty much exclusively.

I know this is an old thread, but I wanted to follow up with my  
recent experience. As stated earlier, I had been running IC-540 in  
rpc mode on Apache 2 with mod_perl and Interchange::Link, and had  
seen the same segfaults as Grant.

A couple weeks ago, Apache began hitting MaxClients and unresponsive  
until restarted. I tinkered with the MaxClients setting and used up  
all my RAM really quick (dumb me). The processes were around 30-40 mb  
with mod_perl IIRC.

Mike suggested getting rid of mod_perl, so I did and went back to CGI  
with vlink and mod_rewrite. The process sizes dropped and I've had no  
problems at all for a week now.

Here's what I am using for sites at the base domain, like  
www.thealertstore.com:

Catalog:
  CGI_URL variable is '/'
  SERVER_NAME and SAMPLEURL are both the base domain

interchange.cfg:
  Catalog line ends with /cgi-bin/shop (FullURL is off)

Apache config is exactly like
  http://www.icdevgroup.org/docs/howtos/apache-rewrite-rules.html
  except for index line:

  RewriteRule ^/(index.(html?)?)?$  /cgi-bin/shop/index  [L,NS,PT]

This line combines the 3 from that howto for the home page. Also,  
adding 'index' on the end of the /cgi-bin/shop/ allows base domains  
to work with "?mv_pc=abc123" etc.

I think that's it. Thanks to everyone who commented on this thread.

--
Josh Lavin
Kingdom Design   http://www.kingdomdesign.com/

_______________________________________________
interchange-users mailing list
suppressed
http://www.icdevgroup.org/mailman/listinfo/interchange-users