[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-06-26 Thread Adrian Cole (JIRA)


[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16523576#comment-16523576
 ] 

Adrian Cole commented on CAMEL-12276:
-

no you can close this. thanks


> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-06-26 Thread Andrea Cosentino (JIRA)


[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16523571#comment-16523571
 ] 

Andrea Cosentino commented on CAMEL-12276:
--

It should be done because spancollector-scribe should be removed.

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-06-26 Thread Claus Ibsen (JIRA)


[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16523566#comment-16523566
 ] 

Claus Ibsen commented on CAMEL-12276:
-

Is there more work here, or has all of this been implemented with the latest 
work on camel-zipkin? eg can we resolve this ticket

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-06-08 Thread Adrian Cole (JIRA)


[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16506561#comment-16506561
 ] 

Adrian Cole commented on CAMEL-12276:
-

[https://github.com/openzipkin/zipkin-reporter-java/pull/113] re-introduces the 
libthrift (scribe) sender which should help library migration. Please do keep 
an eye on removing the transport eventually :)

 

This will be in zipkin-reporter 2.7 releasing today

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-03-05 Thread Adrian Cole (JIRA)

[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16387229#comment-16387229
 ] 

Adrian Cole commented on CAMEL-12276:
-


The last pull request that went in supported the http reporter. So the
main thing is removing the scribe one.


> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-03-05 Thread Claus Ibsen (JIRA)

[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16386056#comment-16386056
 ] 

Claus Ibsen commented on CAMEL-12276:
-

Thanks Adrian

As this is a fair amount of refactor to switch brave to the http reporter and 
whatnot. Do you have some examples we can look at doing that? Or some pointers 
to get on a start of a refactoring?

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
> Fix For: 2.22.0
>
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-02-17 Thread Adrian Cole (JIRA)

[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16368422#comment-16368422
 ] 

Adrian Cole commented on CAMEL-12276:
-

great. the http transport should be the default. There's already the ability to 
plug-in the other two major choices (as well any minor choice like amazon sqs) 
via the zipkin-reporter library. After http, the dominant choice is kafka, and 
quite a ways later rabbitmq.

For this issue, though, the main focus is just removing scribe, as there are 
other issues we should focus time on, such as modeling concerns raised like 
[https://github.com/openzipkin/brave/issues/610|https://github.com/openzipkin/brave/issues/610.]
 As soon as we remove this source of maintenance, all the other activities will 
be more straight-forward.

 

my 2p

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-02-17 Thread Andrea Cosentino (JIRA)

[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16368187#comment-16368187
 ] 

Andrea Cosentino commented on CAMEL-12276:
--

We can try to use a different spancollector.

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)


[jira] [Commented] (CAMEL-12276) Stop requiring scribe transport for zipkin

2018-02-16 Thread Adrian Cole (JIRA)

[ 
https://issues.apache.org/jira/browse/CAMEL-12276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16368084#comment-16368084
 ] 

Adrian Cole commented on CAMEL-12276:
-

discussion here. you'll notice backporting scribe has only been requested by a 
camel maintainer, vs end users which is how we usually prioritize change 
https://github.com/openzipkin/zipkin-reporter-java/issues/95

> Stop requiring scribe transport for zipkin
> --
>
> Key: CAMEL-12276
> URL: https://issues.apache.org/jira/browse/CAMEL-12276
> Project: Camel
>  Issue Type: Improvement
>  Components: camel-zipkin
>Reporter: Adrian Cole
>Priority: Major
>
> Scribe is a long since archived transport originally from Facebook. 
> Unfortunately, when the zipkin camel integration was originally made, it 
> defaulted to the historical scribe transport as opposed to the much wider 
> supported http transport. Scribe is disabled by default on zipkin for a 
> while, due to it pinning to old libraries.
>  
> Due to lack of maintenance here (especially when compared to opentracing), 
> updates around zipkin don't occur organically. In fact http wasn't added 
> until very recently eventhough it predated the camel integration! 
> [https://github.com/apache/camel/pull/2151]
>  
> This lack of keeping up is kindof ok when measured in months, but when 
> measured in years, it has a bad effect. For example, we've been requests 
> (only because of camel) to add scribe into new libraries. It isn't fair to 
> Zipkin to have to add extra burden just because of lack of progression here. 
> We are mostly volunteer run. If we have extra time, it would be better to use 
> that to address more serious concerns, such as the modeling approach or old 
> apis used here.
>  
> Please deprecate and plan to quickly remove the scribe transport. If that 
> isn't a viable option for you, please deprecate and remove your zipkin 
> support, so that we can do it in a more sustainable and more supportable way 
> externally, but without the confusion caused by conflating scribe with 
> zipkin's port.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)