On Jul 13, 2007, at 1:46 AM, Cornelius Jaeger wrote:
Hi Chuck
I did that. Once. Luckily the .com crash came and put that
client out of my misery. :-)
some guys have all the luck
That does sounds like an adaptor problem, possibly with
communicating with wotaskd.
Have you tried to turn on adaptor logging? Windows.... Hmmmm...
I think you create a text file named logWebObjects in C:\Temp
and, later, look for WebObjects.log in that directory. That
should provide some clue as to what specifically is happening.
ok, i've been combing through this
Ooooh. WOAdaptorState. I remember that. Perhaps this is of some
help (also Google for WOAdaptorState)
http://www.wodeveloper.com/omniLists/webobjects-dev/2003/February/
msg00325.html
The rest of that was of no use. You need to capture the part
where it fails.
there seems to be several ways in which the adaptor and wotaskd is
failing,
Did you try that message's suggestion on clearing the shared adaptor
state?
i need some help interpreting these logs though
failure one:
Info: Sending request to instance number 2, port 2028
Info: Trying to contact ASP05Client:2 on (2028)
Info: Caching hostent for screensol-srv11
Info: attempting to connect to SCREENSOL-SRV11.bison-group.com on
port 2028
Info: Created new pooled connection [1] to screensol-srv11:2028
Info: Using pooled connection to screensol-srv11:2028
Info: ASP05Client:2 on (2028) connected [pooled: Yes]
Error: sendbytes(): send error: An existing connection was forcibly
closed by the remote host.
>>>>>>>>>>>>>>>>>>>>>>>was this closed by wotaskd or by my app?
it's a small webservice that is being requested, and when i checked
it it works fine.
That would be your application. I don't recall ever seeing this
before. Do you have limits on the worker thread side and listen
queue set on that application? Is there anything in the app log
(growing numbers of worker threads, exceptions) that might relate to
this?
Does your application have enough memory?
Error: sendbytes(): failed to send data to app
Error: error sending request
Error: Failed to send request
Info: Dumping pooled connection to screensol-srv11(2028)
Info: Marking instance 2 dead
Warn: Marking screensol-srv11:2 unresponsive
That is all normal handling considering the error that happened.
Debug: connectionAttempts = 1, retries = 5
Debug: ac_readConfiguration(): skipped reading config
Info: Attempting failover to new instance at index 14
Debug: Composed URL to '/cgi-bin/WebObjects/ASP05Client.woa/6/wa/
VFClientAction/updateClientPing?
...
Info: Sending request to instance number 6, port 2004
Info: Trying to contact ASP05Client:6 on (2004)
Info: attempting to connect to SCREENSOL-SRV11.bison-group.com on
port 2004
Info: Created new pooled connection [1] to screensol-srv11:2004
Info: Using pooled connection to screensol-srv11:2004
Info: ASP05Client:6 on (2004) connected [pooled: Yes]
Error: sendbytes(): send error: An existing connection was forcibly
closed by the remote host.
Are Apache and the apps on different machines? Could that machine
just be running out of ports or some other resource? The problem
might lie outside of your application. Don't assume that there is
not a problem on the Windows machine itself when under "load".
Error: sendbytes(): failed to send data to app
Error: error sending request
Error: Failed to send request
Info: Dumping pooled connection to screensol-srv11(2004)
Info: Marking instance 6 dead
Warn: Marking screensol-srv11:6 unresponsive
Debug: connectionAttempts = 2, retries = 5
Debug: ac_readConfiguration(): skipped reading config
Info: Attempting failover to new instance at index 12
...
>>>>>>>>>> i double checked the instances it marked dead, and they
are definitely not dead. i can access them via the adaptor and they
work fine, i get the response but for some reason they get marked dead
If the adaptor can't talk to them, they are effectively dead to it,
so it marks them as such. It is doing the right thing based on the
information that it has available.
next failure:
<CGI> new request: /cgi-bin/WebObjects/ASP05Client.woa/wa/
VFClientAction/updateClientPing
Info: <CGI> new request with Query String:
...
Info: Specific instance ASP05Client: not found. Reloading config.
Info: Reading configuration from http://SCREENSOL-SRV11.bison-
group.com:1085/WebObjects/wotaskd.woa/wa/woconfig
Info: attempting to connect to SCREENSOL-SRV11.bison-group.com on
port 1085
Info: Preparing to read config for host: SCREENSOL-SRV11.bison-
group.com
Info: New response: HTTP/1.0 200 Apple WebObjects
Error: Request handling error: The requested application was not
found on this server.
Info: CGI: config url is http://localhost:1085/
Debug: init_adaptor(): beginning initialization. Adaptor version
4.5.1.
Info: init_adaptor(): config options are: String table:
config = http://localhost:1085/
>>>>>>>> no idea what's going on here. it seems that it can't find
the instances. i reconfigured the adaptor to use ip's instead of
the dns name, to no avail.
Yes, it looks like wotaskd is not returning the list of applications
and instances. IIRC, that means a communication problem on the
machine it is running on.
i'm not sure where it's failing, i know the app isn't producing any
errors/exceptions, and that responses are painfully slow, even when
the server has practically no load.
Time the call to super.dispatchRequest() in Application and log it
out. Is that slow? If so, you have a problem in your app (or it is
running out of memory, system resources etc). If that is not slow,
I'd suspect an issue on that machine. You will need to find out
where the time is going. Is it timing out on something?
so i think the requests are getting dropped all the time. when in
direct connect mode on the server the app is really fast.
i think i'll load a vmware image with ubuntu running apache or
something. it's ridiculous that i can't get this to work
(frustration rears it's ugly head)
:-)
# JVMOptions == -Xmx128m -DWOAllowRapidTurnaround=false
i guess that wasn't the problem but i gave it more memory anyway.
how can i access the log files for wotaskd when it's running as a
service. on osx i just changed the output path of *Spawn*.sh?
I don't recall. SpawnOfWotaskd.exe controls it.
thx for your help, i feel like it's getting under control
Much fun to be had here!
yes, it reminds of the that fond time of my youth when i tried to
build a gokart and ended up blendering the neighbors dog in the 2
piston engine.
no end of fun.
=8-0
Chuck
--
Practical WebObjects - for developers who want to increase their
overall knowledge of WebObjects or who are trying to solve specific
problems.
http://www.global-village.net/products/practical_webobjects
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-deploy mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-deploy/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
