You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by Guozhang Wang <gu...@linkedin.com> on 2014/01/17 02:08:15 UTC
Review Request 17006: Patch for KAFKA-1212
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description
-------
KAFKA-1212.v1
Diffs
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Patch for KAFKA-1212
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32382
-----------------------------------------------------------
system_test/utils/kafka_system_test_utils.py
<https://reviews.apache.org/r/17006/#comment61174>
No I think not.
- Guozhang Wang
On Jan. 17, 2014, 1:08 a.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 17, 2014, 1:08 a.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v1
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Patch for KAFKA-1212
Posted by Jun Rao <ju...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32342
-----------------------------------------------------------
system_test/utils/kafka_system_test_utils.py
<https://reviews.apache.org/r/17006/#comment61094>
Are these two sleeps necessary?
- Jun Rao
On Jan. 17, 2014, 1:08 a.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 17, 2014, 1:08 a.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v1
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Jun Rao <ju...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32722
-----------------------------------------------------------
system_test/utils/kafka_system_test_utils.py
<https://reviews.apache.org/r/17006/#comment61671>
So, what we are killing here is the shell script that starts the producer? If so, could we make this clear in the comment?
- Jun Rao
On Jan. 22, 2014, 7:51 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 7:51 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
> On Jan. 27, 2014, 7:01 p.m., Neha Narkhede wrote:
> > system_test/utils/kafka_system_test_utils.py, line 1490
> > <https://reviews.apache.org/r/17006/diff/6/?file=451631#file451631line1490>
> >
> > wouldn't this only kill the producer that is on localhost. But seems like stop_remote_entity()'s goal is to kill the producer on any remote host?
Remotely a producer performance process will be started, which will terminate itself when the number of messages specified has been sent. So we only need to stop the local script which periodically kicks off new producer performance remotely.
- Guozhang
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32857
-----------------------------------------------------------
On Jan. 27, 2014, 6:59 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 27, 2014, 6:59 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Neha Narkhede <ne...@gmail.com>.
> On Jan. 27, 2014, 7:01 p.m., Neha Narkhede wrote:
> > system_test/utils/kafka_system_test_utils.py, line 1490
> > <https://reviews.apache.org/r/17006/diff/6/?file=451631#file451631line1490>
> >
> > wouldn't this only kill the producer that is on localhost. But seems like stop_remote_entity()'s goal is to kill the producer on any remote host?
>
> Guozhang Wang wrote:
> Remotely a producer performance process will be started, which will terminate itself when the number of messages specified has been sent. So we only need to stop the local script which periodically kicks off new producer performance remotely.
I see. Makes sense. Though I was hoping to see a more generic "stop_local_processes()" and "stop_remote_processes()" APIs.
- Neha
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32857
-----------------------------------------------------------
On Jan. 28, 2014, 5:26 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 28, 2014, 5:26 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.4
>
>
> KAFKA-1212.v3.3
>
>
> KAFKA-1212.v3.2
>
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Neha Narkhede <ne...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32857
-----------------------------------------------------------
system_test/utils/kafka_system_test_utils.py
<https://reviews.apache.org/r/17006/#comment61858>
wouldn't this only kill the producer that is on localhost. But seems like stop_remote_entity()'s goal is to kill the producer on any remote host?
- Neha Narkhede
On Jan. 27, 2014, 6:59 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 27, 2014, 6:59 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Neha Narkhede <ne...@gmail.com>.
> On Feb. 13, 2014, 10:40 p.m., Joel Koshy wrote:
> > system_test/utils/kafka_system_test_utils.py, line 937
> > <https://reviews.apache.org/r/17006/diff/9/?file=452899#file452899line937>
> >
> > This does not seem right (or I may be misunderstanding the issue in this
> > jira).
> >
> > This will return the PID of the locally running test-runner. You will need
> > to parse out the PPID from the response of the ssh "start and echo pid"
> > command - see kafka_system_test_utils.start_entity_in_background for how the
> > PPID is obtained for the broker/console consumer, etc.
> >
> >
Guozhang's reply to the same question I asked as well :)
"Remotely a producer performance process will be started, which will terminate itself when the number of messages specified has been sent. So we only need to stop the local script which periodically kicks off new producer performance remotely."
- Neha
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review34432
-----------------------------------------------------------
On Jan. 28, 2014, 5:26 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 28, 2014, 5:26 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.4
>
>
> KAFKA-1212.v3.3
>
>
> KAFKA-1212.v3.2
>
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Guozhang Wang <gu...@linkedin.com>.
> On Feb. 13, 2014, 10:40 p.m., Joel Koshy wrote:
> > system_test/utils/kafka_system_test_utils.py, line 937
> > <https://reviews.apache.org/r/17006/diff/9/?file=452899#file452899line937>
> >
> > This does not seem right (or I may be misunderstanding the issue in this
> > jira).
> >
> > This will return the PID of the locally running test-runner. You will need
> > to parse out the PPID from the response of the ssh "start and echo pid"
> > command - see kafka_system_test_utils.start_entity_in_background for how the
> > PPID is obtained for the broker/console consumer, etc.
> >
> >
>
> Neha Narkhede wrote:
> Guozhang's reply to the same question I asked as well :)
>
> "Remotely a producer performance process will be started, which will terminate itself when the number of messages specified has been sent. So we only need to stop the local script which periodically kicks off new producer performance remotely."
>
>
> Joel Koshy wrote:
> I think the loop that spawns off remote producers will end when the "stopBackgroundProducer" flag is set to true (which is already the case). What seems to happen with this patch is the local pid (of the test runner is taken) and in stop remote entities, we will ssh to that remote entity and SIGTERM a pid that doesn't even exist on that remote entity. i.e., I don't think this actually kills the local script. Furthermore in this case I do think we should SIGTERM the remote producer and not wait for them to terminate themselves.
The issue here is that if one test case ends abnormally, the script will not be terminated, and hence will keep spawning remote producers and hence mess up with the subsequent test cases. What this patch does it that in stop_all_remote_running_processes, which is called at the end of each test case no matter if it exists normally or not, check if the local script still exists. If yes, signal it to terminate. The PID value is actually not used, we only check the length of the pid list. Does this make more sense now?
- Guozhang
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review34432
-----------------------------------------------------------
On Jan. 28, 2014, 5:26 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 28, 2014, 5:26 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.4
>
>
> KAFKA-1212.v3.3
>
>
> KAFKA-1212.v3.2
>
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Joel Koshy <jj...@gmail.com>.
> On Feb. 13, 2014, 10:40 p.m., Joel Koshy wrote:
> > system_test/utils/kafka_system_test_utils.py, line 937
> > <https://reviews.apache.org/r/17006/diff/9/?file=452899#file452899line937>
> >
> > This does not seem right (or I may be misunderstanding the issue in this
> > jira).
> >
> > This will return the PID of the locally running test-runner. You will need
> > to parse out the PPID from the response of the ssh "start and echo pid"
> > command - see kafka_system_test_utils.start_entity_in_background for how the
> > PPID is obtained for the broker/console consumer, etc.
> >
> >
>
> Neha Narkhede wrote:
> Guozhang's reply to the same question I asked as well :)
>
> "Remotely a producer performance process will be started, which will terminate itself when the number of messages specified has been sent. So we only need to stop the local script which periodically kicks off new producer performance remotely."
>
I think the loop that spawns off remote producers will end when the "stopBackgroundProducer" flag is set to true (which is already the case). What seems to happen with this patch is the local pid (of the test runner is taken) and in stop remote entities, we will ssh to that remote entity and SIGTERM a pid that doesn't even exist on that remote entity. i.e., I don't think this actually kills the local script. Furthermore in this case I do think we should SIGTERM the remote producer and not wait for them to terminate themselves.
- Joel
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review34432
-----------------------------------------------------------
On Jan. 28, 2014, 5:26 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 28, 2014, 5:26 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.4
>
>
> KAFKA-1212.v3.3
>
>
> KAFKA-1212.v3.2
>
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Joel Koshy <jj...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review34432
-----------------------------------------------------------
system_test/utils/kafka_system_test_utils.py
<https://reviews.apache.org/r/17006/#comment64523>
This does not seem right (or I may be misunderstanding the issue in this
jira).
This will return the PID of the locally running test-runner. You will need
to parse out the PPID from the response of the ssh "start and echo pid"
command - see kafka_system_test_utils.start_entity_in_background for how the
PPID is obtained for the broker/console consumer, etc.
- Joel Koshy
On Jan. 28, 2014, 5:26 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 28, 2014, 5:26 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v3.4
>
>
> KAFKA-1212.v3.3
>
>
> KAFKA-1212.v3.2
>
>
> KAFKA-1212.v3.1
>
>
> KAFKA-1212.v3
>
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Check if there are hanging producer threads in
stopping all entities
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 28, 2014, 5:26 p.m.)
Review request for kafka.
Summary (updated)
-----------------
Check if there are hanging producer threads in stopping all entities
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v3.4
KAFKA-1212.v3.3
KAFKA-1212.v3.2
KAFKA-1212.v3.1
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 28, 2014, 2:55 a.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v3.3
KAFKA-1212.v3.2
KAFKA-1212.v3.1
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 28, 2014, 2:24 a.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v3.2
KAFKA-1212.v3.1
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 27, 2014, 6:59 p.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v3.1
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 27, 2014, 6:55 p.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description
-------
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 24, 2014, 7:44 p.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v3
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
> On Jan. 23, 2014, 5:51 p.m., John Fung wrote:
> > There will be only 1 producer_performance running per entity_id. Therefore, entity_id can be used as a key in the dictionary to keep track of multiple running producer_performance threads:
> >
> > + os.tcaseEnv.producerHostParentPidDict[entityId] = os.getpid()
Hi John, thanks for the comments. Not sure I get it clear though. What do you suggest to change the above line?
- Guozhang
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32632
-----------------------------------------------------------
On Jan. 22, 2014, 7:51 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 7:51 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by John Fung <jo...@gmail.com>.
> On Jan. 23, 2014, 5:51 p.m., John Fung wrote:
> > There will be only 1 producer_performance running per entity_id. Therefore, entity_id can be used as a key in the dictionary to keep track of multiple running producer_performance threads:
> >
> > + os.tcaseEnv.producerHostParentPidDict[entityId] = os.getpid()
>
> Guozhang Wang wrote:
> Hi John, thanks for the comments. Not sure I get it clear though. What do you suggest to change the above line?
Guozhang, sorry about the confusion. I meant using entityId as a key to the dictionary is the proper way to track producer pids. I wanted to give a "+1" for it.
- John
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32632
-----------------------------------------------------------
On Jan. 22, 2014, 7:51 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 7:51 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by John Fung <jo...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/#review32632
-----------------------------------------------------------
There will be only 1 producer_performance running per entity_id. Therefore, entity_id can be used as a key in the dictionary to keep track of multiple running producer_performance threads:
+ os.tcaseEnv.producerHostParentPidDict[entityId] = os.getpid()
- John Fung
On Jan. 22, 2014, 7:51 p.m., Guozhang Wang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/17006/
> -----------------------------------------------------------
>
> (Updated Jan. 22, 2014, 7:51 p.m.)
>
>
> Review request for kafka.
>
>
> Bugs: KAFKA-1212
> https://issues.apache.org/jira/browse/KAFKA-1212
>
>
> Repository: kafka
>
>
> Description
> -------
>
> KAFKA-1212.v2
>
>
> Diffs
> -----
>
> system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
>
> Diff: https://reviews.apache.org/r/17006/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Guozhang Wang
>
>
Re: Review Request 17006: Keep track of local pid and kill it on stopping
all entities as discussed with John
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 22, 2014, 7:51 p.m.)
Review request for kafka.
Summary (updated)
-----------------
Keep track of local pid and kill it on stopping all entities as discussed with John
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v2
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang
Re: Review Request 17006: Patch for KAFKA-1212
Posted by Guozhang Wang <gu...@linkedin.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/17006/
-----------------------------------------------------------
(Updated Jan. 21, 2014, 9:21 p.m.)
Review request for kafka.
Bugs: KAFKA-1212
https://issues.apache.org/jira/browse/KAFKA-1212
Repository: kafka
Description (updated)
-------
KAFKA-1212.v1.1
KAFKA-1212.v1
Diffs (updated)
-----
system_test/utils/kafka_system_test_utils.py fb4a9c05bf6f39a7abf41126325ed5ca26bcc246
Diff: https://reviews.apache.org/r/17006/diff/
Testing
-------
Thanks,
Guozhang Wang