rave-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gustavo Monarin <gustavomona...@gmail.com>
Subject Re: opensocial makeRequest time out
Date Wed, 04 Dec 2013 17:03:00 GMT
Hi Stanton,

Thanks a bunch for so detailed answer.  I was not sure if here was the
proper list to ask this.

I will try this and i was  also thinking to replace all the makeRequest
from gadgest.io by some xhr library implementation directly, in my case by
the jquery.get..

Regards,


On Wed, Dec 4, 2013 at 3:36 PM, Stanton Sievers <ssievers@apache.org> wrote:

> Hi Gustavo,
>
> Your issue is likely that the socket is timing out and not the connection.
> The BasicHttpFetcher treats any number of timeout exceptions in the same
> way (ConnectionPoolTimeoutException, SocketTimeoutException,
> SocketException, HttpHostConnectException.... and so on).
>
> Try setting "shindig.http.client.read-timeout-ms" as well.  This may
> default to 0 (infinite) but I'm not sure off the top of my head.
>
> If that doesn't work, you should ensure that you are setting the timeouts
> in the correct place.  You can be very certain that the values are being
> used by setting one of the timeouts to 0 or less.  An exception will be
> thrown in this case.
>
> Also, I've written a Shindig bug [1] to help diagnose these problems in the
> future.
>
> Thanks,
> -Stanton
>
> [1] https://issues.apache.org/jira/browse/SHINDIG-1961
>
>
>
>
> On Sun, Dec 1, 2013 at 3:44 PM, Gustavo Monarin <gustavomonarin@gmail.com
> >wrote:
>
> > Hi all,
> >
> > I have been running apache rave in a custom portal and i am experiencing
> > some time-out problems with shindig using gadgets.io.makeRequest.
> >
> > In the logs there is the following warning:
> >
> > Dec 01, 2013 9:33:11 PM org.apache.shindig.gadgets.http.BasicHttpFetcher
> > fetch
> > INFO: http://localhost:8080 .....  has timed out because of the
> following
> > exception: org.apache.shindig.gadgets.http.BasicHttpFetcher - Read timed
> > out - 5,008 ms.
> >
> > This happens when a widget tries to request a report in another context
> in
> > the same container. In the logs you can see that the report is still
> being
> > processed.
> >
> > I already change the rave.shindig.properties related to the
> BasicHttpFecher
> > to 30s but still gets the 5s timeout:
> > # Configuration for the HttpFetcher
> > # Connection timeout, in milliseconds, for requests.
> > shindig.http.client.connection-timeout-ms=30000
> >
> > I have ensured that the properties file is being used by changing the
> > datasource name to an invalid name and checked the exceptions.
> >
> > Any suggestions?
> >
> > Should i post it directly in the shindig project ?
> >
> > Thanks in advance.
> >
> > --
> >
> > Gustavo Monarin de Sousa
> >
>



-- 

Gustavo Monarin de Sousa

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message