[LON-CAPA-admin] Students not seeing homework on access servers

[Stuart Raeburn]

Richard,

> Apache2::RequestIO::print: (103) Software caused connection abort at....
>
> And then there are different files they are getting that at
>
> at /home/httpd/lib/perl//Apache/lonerrorhandler.pm line 53
> at /home/httpd/lib/perl//Apache/lonnavmaps.pm line 1831
> at /home/httpd/lib/perl//Apache/lonnavmaps.pm line 1864
> at /home/httpd/lib/perl//Apache/lonnavdisplay.pm line 118
> at /home/httpd/lib/perl//Apache/lonhomework.pm line 1108
> and others.

This usually means a student got tired of waiting for the page to  
load, and clicked away.  As a result the server's response could be  
longer be returned to the client, so the message "Software caused  
connection abort" is logged and Apache/mod_perl aborts at the current  
$r->print() line, which is what you'll find at those lines in each of  
the reported perl modules.

The first of these -- the one involving lonerrorhandler.pm is  
potentially the most interesting, as it suggests an internal server  
error occurred first, so you might look in the Apache error_log to  
find out what the error was.

Of course, students clicking away, is a symptom.  The larger question  
is why are pages loading slowly on your access servers.  The first  
thing I would check is that the memcached daemon is running:

ps -ef |grep memcached |grep -v grep

(The output should be something like:
www  <process> 1  0 <date> <time> /usr/bin/memcached -m 400 -v

and is caching; see my post to the admin list from earlier this year  
for a script you can use to dump the contents of memcached:

http://mail.lon-capa.org/pipermail/lon-capa-admin/2015-February/003005.html

Also see:
http://mail.lon-capa.org/pipermail/lon-capa-admin/2015-February/003013.html


Stuart Raeburn
LON-CAPA Academic Consortium


Quoting "Lucas, Richard" <clucas at purdue.edu>:

> Ok apparently the student issue was a false alarm raised by the   
> instructor so no issues there. I am still wondering if there is   
> anything I can do about the errors though.
>
> Thanks,
> Richard
>
> From: lon-capa-admin-bounces at mail.lon-capa.org   
> [mailto:lon-capa-admin-bounces at mail.lon-capa.org] On Behalf Of   
> Lucas, Richard
> Sent: Sunday, September 20, 2015 4:04 PM
> To: lon-capa-admin at mail.lon-capa.org
> Subject: [LON-CAPA-admin] Students not seeing homework on access servers
>
> We just added two access servers on Friday and today there are many   
> students reporting they are unable to see their homework. I am not   
> sure what to look at here, but I do see a bunch of errors in the   
> httpd error_log. Most of them start with this:
>
> Apache2::RequestIO::print: (103) Software caused connection abort at....
>
> And then there are different files they are getting that at
>
> at /home/httpd/lib/perl//Apache/lonerrorhandler.pm line 53
> at /home/httpd/lib/perl//Apache/lonnavmaps.pm line 1831
> at /home/httpd/lib/perl//Apache/lonnavmaps.pm line 1864
> at /home/httpd/lib/perl//Apache/lonnavdisplay.pm line 118
> at /home/httpd/lib/perl//Apache/lonhomework.pm line 1108
> and others.
>
> I also see some other messages like this, but they are a lot fewer:
>
> [Sun Sep 20 15:21:09 2015] [error] access to   
> /res/purdue/purdue_math/math15800/Polynomial and Rational   
> Functions/Rational Functions/Rational   
> Functions_problems/asyp1_03_d.problem faile
> d for 10.192.57.242, reason: Invalid symb for   
> /res/purdue/purdue_math/math15800/Polynomial and Rational   
> Functions/Rational Functions/Rational   
> Functions_problems/asyp1_03_d.problem: upl
> oaded/purdue/4o12229eff3c955c9purduel1/default_1442167687.sequence___4___purdue/purdue_math/math15800/Polynomial and Rational Functions/Rational Functions/Rational   
> Functions_problems/a
> syp1_03_d.problem
> [Sun Sep 20 15:21:09 2015] [error] access to   
> /res/purdue/purdue_math/math15800/Polynomial and Rational   
> Functions/Rational Functions/Rational   
> Functions_problems/asyp1_03_d.problem faile
> d for 10.192.57.242, reason: Invalid Access for peng119 domain   
> purdue access bre
>
> Any ideas here?
>
>
> Thanks,
> Richard