You are viewing a plain text version of this content. The canonical link for it is here.
Posted to ivy-user@ant.apache.org by jgunz <sl...@twcny.rr.com> on 2007/04/10 16:46:43 UTC
Sporadic resolution errors
I'm having an issue where Ivy sporadically fails to resolve dependencies that
actually DO exist in my internal Ivy repository. I believe the problem is
caused by heavy processor or network load on my development machines because
usually rerunning the build will succeed. Is there any configuration I could
do to perhaps extend the timeout used or add a retry attempt to resolution?
An example failure I got this morning is...
[FAILED ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar] : Socket is closed
(46ms)
[NOT FOUND ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar]
Antlr does exist, when I reran my build it succeeded. The "socket is closed"
message is not always the same -- I've seen a variety of error messages in
this spot.
Any help would be appreciated.
--
View this message in context: http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921261
Sent from the ivy-user mailing list archive at Nabble.com.
RE: Sporadic resolution errors
Posted by jgunz <sl...@twcny.rr.com>.
I'll see if I can make some time to implement one of the proposed solutions.
If I do, I'll let you know. Thanks for the help and the great product.
--
View this message in context: http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9941566
Sent from the ivy-user mailing list archive at Nabble.com.
RE: Sporadic resolution errors
Posted by Gilles Scokart <gs...@gmail.com>.
Another workaround might be to reduce the load on your repository by placing
a squid proxy in front or your repository, or maybe look at some options in
the apache server that might enhance the caching. I never did such a thing,
but it might work...
However, if you don't know how to configure apache orsquid, and if you are a
java developer, writing a resolver might be better (certainly for the
community :-)).
Gilles
> -----Original Message-----
> From: Xavier Hanin [mailto:xavier.hanin@gmail.com]
> Sent: mardi 10 avril 2007 18:20
> To: ivy-user@incubator.apache.org
> Subject: Re: Sporadic resolution errors
>
> On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
> >
> >
> > We are using the url resolver to access a Subversion controlled Ivy
> > repository via https. Both the build machine and the machine hosting the
> > repository have been under excessive loads lately and I suspect this is
> > the
> > problem -- I would just like to avoid having our Cruise Control process
> > fail
> > and upsetting the rest of the team ;-)
>
>
> I understand, but for the moment the only thing you can do is either patch
> Ivy to support multiple attempts and customizable timeouts (and contribute
> the patch to the community :-)) or write your own resolver (or even simply
> repository) implementing this behavior on top of an existing repository
> implementation (like URLRepository) or by delegation (which would make
> your
> implemenation usable with any repository impl). This could also be
> contributed to the community. The last solution I see is to open a JIRA
> issue and wait for someone in the community to consider it.
>
> - Xavier
>
> Xavier Hanin wrote:
> > >
> > > On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
> > >>
> > >>
> > >> I'm having an issue where Ivy sporadically fails to resolve
> > dependencies
> > >> that
> > >> actually DO exist in my internal Ivy repository. I believe the
> problem
> > is
> > >> caused by heavy processor or network load on my development machines
> > >> because
> > >> usually rerunning the build will succeed. Is there any configuration
> I
> > >> could
> > >> do to perhaps extend the timeout used or add a retry attempt to
> > >> resolution?
> > >>
> > >> An example failure I got this morning is...
> > >>
> > >> [FAILED ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar] : Socket is
> > >> closed
> > >> (46ms)
> > >> [NOT FOUND ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar]
> > >>
> > >>
> > >> Antlr does exist, when I reran my build it succeeded. The "socket is
> > >> closed"
> > >> message is not always the same -- I've seen a variety of error
> messages
> > >> in
> > >> this spot.
> > >
> > >
> > > Never seen that before. Which resolver are you using?
> > > For the timeout or retry configuration, unfortunately there is no such
> > > feature. You can add a JIRA issue for that.
> > >
> > > - Xavier
> > >
> > > Any help would be appreciated.
> > >> --
> > >> View this message in context:
> > >>
> > http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921261
> > >> Sent from the ivy-user mailing list archive at Nabble.com.
> > >>
> > >>
> > >
> > >
> > > --
> > > Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
> > > Manage your dependencies with Ivy!
> > > http://incubator.apache.org/ivy/
> > >
> > >
> >
> > --
> > View this message in context:
> > http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921796
> > Sent from the ivy-user mailing list archive at Nabble.com.
> >
> >
>
>
> --
> Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
> Manage your dependencies with Ivy!
> http://incubator.apache.org/ivy/
Re: Sporadic resolution errors
Posted by Xavier Hanin <xa...@gmail.com>.
On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
>
>
> We are using the url resolver to access a Subversion controlled Ivy
> repository via https. Both the build machine and the machine hosting the
> repository have been under excessive loads lately and I suspect this is
> the
> problem -- I would just like to avoid having our Cruise Control process
> fail
> and upsetting the rest of the team ;-)
I understand, but for the moment the only thing you can do is either patch
Ivy to support multiple attempts and customizable timeouts (and contribute
the patch to the community :-)) or write your own resolver (or even simply
repository) implementing this behavior on top of an existing repository
implementation (like URLRepository) or by delegation (which would make your
implemenation usable with any repository impl). This could also be
contributed to the community. The last solution I see is to open a JIRA
issue and wait for someone in the community to consider it.
- Xavier
Xavier Hanin wrote:
> >
> > On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
> >>
> >>
> >> I'm having an issue where Ivy sporadically fails to resolve
> dependencies
> >> that
> >> actually DO exist in my internal Ivy repository. I believe the problem
> is
> >> caused by heavy processor or network load on my development machines
> >> because
> >> usually rerunning the build will succeed. Is there any configuration I
> >> could
> >> do to perhaps extend the timeout used or add a retry attempt to
> >> resolution?
> >>
> >> An example failure I got this morning is...
> >>
> >> [FAILED ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar] : Socket is
> >> closed
> >> (46ms)
> >> [NOT FOUND ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar]
> >>
> >>
> >> Antlr does exist, when I reran my build it succeeded. The "socket is
> >> closed"
> >> message is not always the same -- I've seen a variety of error messages
> >> in
> >> this spot.
> >
> >
> > Never seen that before. Which resolver are you using?
> > For the timeout or retry configuration, unfortunately there is no such
> > feature. You can add a JIRA issue for that.
> >
> > - Xavier
> >
> > Any help would be appreciated.
> >> --
> >> View this message in context:
> >>
> http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921261
> >> Sent from the ivy-user mailing list archive at Nabble.com.
> >>
> >>
> >
> >
> > --
> > Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
> > Manage your dependencies with Ivy!
> > http://incubator.apache.org/ivy/
> >
> >
>
> --
> View this message in context:
> http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921796
> Sent from the ivy-user mailing list archive at Nabble.com.
>
>
--
Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
Manage your dependencies with Ivy!
http://incubator.apache.org/ivy/
Re: Sporadic resolution errors
Posted by jgunz <sl...@twcny.rr.com>.
We are using the url resolver to access a Subversion controlled Ivy
repository via https. Both the build machine and the machine hosting the
repository have been under excessive loads lately and I suspect this is the
problem -- I would just like to avoid having our Cruise Control process fail
and upsetting the rest of the team ;-)
Xavier Hanin wrote:
>
> On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
>>
>>
>> I'm having an issue where Ivy sporadically fails to resolve dependencies
>> that
>> actually DO exist in my internal Ivy repository. I believe the problem is
>> caused by heavy processor or network load on my development machines
>> because
>> usually rerunning the build will succeed. Is there any configuration I
>> could
>> do to perhaps extend the timeout used or add a retry attempt to
>> resolution?
>>
>> An example failure I got this morning is...
>>
>> [FAILED ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar] : Socket is
>> closed
>> (46ms)
>> [NOT FOUND ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar]
>>
>>
>> Antlr does exist, when I reran my build it succeeded. The "socket is
>> closed"
>> message is not always the same -- I've seen a variety of error messages
>> in
>> this spot.
>
>
> Never seen that before. Which resolver are you using?
> For the timeout or retry configuration, unfortunately there is no such
> feature. You can add a JIRA issue for that.
>
> - Xavier
>
> Any help would be appreciated.
>> --
>> View this message in context:
>> http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921261
>> Sent from the ivy-user mailing list archive at Nabble.com.
>>
>>
>
>
> --
> Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
> Manage your dependencies with Ivy!
> http://incubator.apache.org/ivy/
>
>
--
View this message in context: http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921796
Sent from the ivy-user mailing list archive at Nabble.com.
Re: Sporadic resolution errors
Posted by Xavier Hanin <xa...@gmail.com>.
On 4/10/07, jgunz <sl...@twcny.rr.com> wrote:
>
>
> I'm having an issue where Ivy sporadically fails to resolve dependencies
> that
> actually DO exist in my internal Ivy repository. I believe the problem is
> caused by heavy processor or network load on my development machines
> because
> usually rerunning the build will succeed. Is there any configuration I
> could
> do to perhaps extend the timeout used or add a retry attempt to
> resolution?
>
> An example failure I got this morning is...
>
> [FAILED ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar] : Socket is
> closed
> (46ms)
> [NOT FOUND ] [ antlr | antlr | 2.7.6rc1 ]/antlr.jar[jar]
>
>
> Antlr does exist, when I reran my build it succeeded. The "socket is
> closed"
> message is not always the same -- I've seen a variety of error messages in
> this spot.
Never seen that before. Which resolver are you using?
For the timeout or retry configuration, unfortunately there is no such
feature. You can add a JIRA issue for that.
- Xavier
Any help would be appreciated.
> --
> View this message in context:
> http://www.nabble.com/Sporadic-resolution-errors-tf3553442.html#a9921261
> Sent from the ivy-user mailing list archive at Nabble.com.
>
>
--
Learn Ivy at ApacheCon: http://www.eu.apachecon.com/
Manage your dependencies with Ivy!
http://incubator.apache.org/ivy/