Discussion list for the Yaws web server, a high perfomance http 1.1 server written in Erlang.

Roberto Saccon | 1 Dec 2006 04:52

JSON-RPC problems with latest dojo toolkit

I updated my dojo Javascript framework to trunk (and also tested with
0.4.1-RC2) I now I get this with any JSON RPC call (and I get no HTPP
response at all):

=ERROR REPORT==== 1-Dec-2006::00:43:06 ===
Yaws process died: {function_clause,[{yaws_server,
                                         binary_size,
                                         [3409,{status,200}]},
                                     {yaws_server,binary_size,2},
                                     {yaws_server,deliver_accumulated,5},
                                     {yaws_server,finish_up_dyn_file,2},
                                     {yaws_server,aloop,3},
                                     {yaws_server,acceptor0,2},
                                     {proc_lib,init_p,5}]}

=ERROR REPORT==== 1-Dec-2006::00:43:06 ===
Yaws process died: {function_clause,[{yaws_server,
                                         binary_size,
                                         [3409,{status,200}]},
                                     {yaws_server,binary_size,2},
                                     {yaws_server,deliver_accumulated,5},
                                     {yaws_server,finish_up_dyn_file,2},
                                     {yaws_server,aloop,3},
                                     {yaws_server,acceptor0,2},
                                     {proc_lib,init_p,5}]}

What does that mean ? And yes, I get it twice per HTTP JSON request. I
am using yaws 1.65, Erlang R11B-0 and Ubutu 6.10. If I do the JSON-RPC
test provided with yaws and the jsolait toolkit, everything works as
expected.

(Continue reading)

headers

Mikael Magnusson | 1 Dec 2006 15:43

"compile" application dependency

Hi,

I think the "compile" application should be added as a dependency to 
yaws.app. Without the compile application yaws will hang when accessing 
a .yaws page.

Regards,
Mikael

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Claes Wikström | 1 Dec 2006 15:51

Re: "compile" application dependency

Mikael Magnusson wrote:
> Hi,
> 
> I think the "compile" application should be added as a dependency to 
> yaws.app. Without the compile application yaws will hang when accessing 
> a .yaws page.
> 

Nahhh ... lot's of people use yaws entirely without any .yaws
pages. One common setup is static pages gifs etc + one
fatso appmod which do AJAX stuff. No .yaws pages at all.
Requiring compiler will then ( I guess ??) make application:load(yaws)
fail if compiler isn't there.

/klacke

--

-- 
Claes Wikstrom                        -- Caps lock is nowhere and
http://www.tail-f.com                 -- everything is under control
cellphone: +46 70 2097763

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Count László de Almásy | 3 Dec 2006 19:57

"TCP: Treason uncloaked" messages with Linux/Yaws

Since switching from Apache to Yaws last week I'm now seeing some
messages in my kernel log that look like this:

TCP: Treason uncloaked! Peer 84.254.135.6:3936/80 shrinks window
342474125:342474126. Repaired.
TCP: Treason uncloaked! Peer 84.254.135.6:4058/80 shrinks window
1648658911:1648658912. Repaired.
TCP: Treason uncloaked! Peer 84.254.135.6:4062/80 shrinks window
1668582427:1668582428. Repaired.
TCP: Treason uncloaked! Peer 84.254.135.6:2580/80 shrinks window
3941448823:3941448824. Repaired.
TCP: Treason uncloaked! Peer 84.254.135.6:2680/80 shrinks window
2056818233:2056818234. Repaired.

This is on Linux 2.6.17.9.  Anyone seen the same?  And why would this
only be an issue with Yaws and not Apache?

--

-- 
Cheers, László

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Kirill Zaborski | 3 Dec 2006 20:28

index.yaws disappears

I start YAWS in embedded mode and my index.yaws which is situated in my docroot folder disappears everytime.
The code I use :

    GL=[{trace,false},{tmpdir,"P:/tmp"},{flags,[{auth_log,false},{copy_errlog,false}]}],
    SL=[{port,777},{servername,"myserver"},{listen,{0,0,0,0}},{dir_listings, true},{flags,[{access_log,false}]}],
    yaws:start_embedded("p:/tmp/yaws", SL,GL).

What am I doing wrong?

BR,
Kirill.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

_______________________________________________
Erlyaws-list mailing list
Erlyaws-list@...
https://lists.sourceforge.net/lists/listinfo/erlyaws-list

headers

Claes Wikström | 4 Dec 2006 09:48

Re: index.yaws disappears

Kirill Zaborski wrote:
> I start YAWS in embedded mode and my index.yaws which is situated in my 
> docroot folder disappears everytime.

This was a new one 

I have no idea who removes the index.yaws file.
For a webserver to remove top index file would be a very odd thing to do.....
to say the least. Needs to be debugged.

/klacke

--

-- 
Claes Wikstrom                        -- Caps lock is nowhere and
http://www.tail-f.com                 -- everything is under control
cellphone: +46 70 2097763

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Mikael Karlsson | 5 Dec 2006 20:29

yaws_server:gserv_loop/1 goes into spin

I have a problem when doing a yaws_api:setconf in R11B on sparc_solaris. I do 
not know the root cause  yet but I can see that the funs 
yaws_server:gserv_loop and acceptor0 goes into a infinite loop when the 
do_accept in acceptor0 returns with an error ({error,einval}).

gserv_loop enters the clause:

 	{_From, next} when Ready == [] ->
	    New = acceptor(GS),
	    gserv_loop(GS, Ready, Rnum, New);

the call to acceptor spawns a process with fun acceptor0, which first calls 
do_accept and then sends {self(), next} back to the gserv_loop which calls 
acceptor again, and so forth. Since do_accept returns immediately with an 
error everything goes into a spin. 

acceptor0(GS, Top) ->
    ?TC([{record, GS, gs}]),
    put(gc, GS#gs.gconf),
    X = do_accept(GS),
    Top ! {self(), next},

I guess that there should be another handling of this when do_accept returns 
with an error.

Regards
Mikael

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Kirill Zaborski | 5 Dec 2006 23:44

Re: index.yaws disappears

Now I have resolved the issues I had with CEAN (it seems so), so tomorrow I'll try to find the one who is responsible for this "disappearance".
Have you read my email about the bug in YAWS on erlang-questions?

BR,
Kirill.

On 12/4/06, Claes Wikström <klacke-RbUkv7OeYy7QT0dZR+AlfA@public.gmane.org> wrote:

Kirill Zaborski wrote:
> I start YAWS in embedded mode and my index.yaws which is situated in my
> docroot folder disappears everytime.

This was a new one

I have no idea who removes the index.yaws file.
For a webserver to remove top index file would be a very odd thing to do.....
to say the least. Needs to be debugged.

/klacke

--
Claes Wikstrom -- Caps lock is nowhere and
http://www.tail-f.com -- everything is under control
cellphone: +46 70 2097763

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

_______________________________________________
Erlyaws-list mailing list
Erlyaws-list@...
https://lists.sourceforge.net/lists/listinfo/erlyaws-list

headers

Claes Wikstrom | 5 Dec 2006 23:55

Re: yaws_server:gserv_loop/1 goes into spindo_acc

Mikael Karlsson wrote:
> I have a problem when doing a yaws_api:setconf in R11B on sparc_solaris. I do 
> not know the root cause  yet but I can see that the funs 
> yaws_server:gserv_loop and acceptor0 goes into a infinite loop when the 
> do_accept in acceptor0 returns with an error ({error,einval}).
> 
> gserv_loop enters the clause:
> 
>  	{_From, next} when Ready == [] ->
> 	    New = acceptor(GS),
> 	    gserv_loop(GS, Ready, Rnum, New);
> 
> the call to acceptor spawns a process with fun acceptor0, which first calls 
> do_accept and then sends {self(), next} back to the gserv_loop which calls 
> acceptor again, and so forth. Since do_accept returns immediately with an 
> error everything goes into a spin. 
> 
> acceptor0(GS, Top) ->
>     ?TC([{record, GS, gs}]),
>     put(gc, GS#gs.gconf),
>     X = do_accept(GS),
>     Top ! {self(), next},
> 
> I guess that there should be another handling of this when do_accept returns 
> with an error.

Hmmm, the (TM) right behaviour if accept fails is ... probably
to try to accept() again. Anybody's got an opinion on that ?

In your case, I think the loop stems from
fault embedded mode configs. einval is the reason, the args to accept
are somehow weird.

You need to debug this, stage 1 would be to print out the
arg to accept()

/klacke

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

headers

Claes Wikstrom | 6 Dec 2006 00:00

Re: index.yaws disappears

Kirill Zaborski wrote:
> Now I have resolved the issues I had with CEAN (it seems so), so 
> tomorrow I'll try to find the one who is responsible for this 
> "disappearance".
> Have you read my email about the bug in YAWS on erlang-questions?

I did now, I'll have a look at it .. thanks

/klacke

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

5	May 2013
11	April 2013
25	March 2013
5	February 2013
2	January 2013
14	December 2012
25	November 2012
32	October 2012
3	September 2012
21	August 2012
13	July 2012
42	June 2012
11	May 2012
28	April 2012
29	March 2012
46	February 2012
25	January 2012
16	December 2011
10	November 2011
34	October 2011
31	September 2011
5	August 2011
25	July 2011
55	June 2011
61	May 2011
52	April 2011
43	March 2011
20	February 2011
18	January 2011
20	December 2010
7	November 2010
31	October 2010
50	September 2010
58	August 2010
47	July 2010
19	June 2010
54	May 2010
47	April 2010
40	March 2010
43	February 2010
30	January 2010
74	December 2009
28	November 2009
32	October 2009
50	September 2009
51	August 2009
72	July 2009
40	June 2009
10	May 2009
35	April 2009
40	March 2009
66	February 2009
40	January 2009
51	December 2008
51	November 2008
21	October 2008
28	September 2008
45	August 2008
20	July 2008
28	June 2008
42	May 2008
51	April 2008
24	March 2008
26	February 2008
27	January 2008
50	December 2007
38	November 2007
38	October 2007
46	September 2007
4	August 2007
2	July 2007
12	June 2007
28	May 2007
63	April 2007
80	March 2007
71	February 2007
61	January 2007
32	December 2006
98	November 2006
38	October 2006
103	September 2006
112	August 2006
42	July 2006
84	June 2006
11	May 2006
29	April 2006
17	March 2006
39	February 2006
4	January 2006
31	December 2005
34	November 2005
6	October 2005
16	September 2005
9	August 2005
14	July 2005
18	June 2005
42	May 2005
57	April 2005
21	March 2005
12	February 2005
14	January 2005
28	December 2004

Mon	Tue	Wed	Thu	Fri	Sat	Sun

				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28	29	30	31