Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/
---

(Updated April 21, 2015, 7:18 p.m.)


Review request for mesos and Ben Mahler.


Changes
---

comments. NNFR.


Bugs: MESOS-1127
https://issues.apache.org/jira/browse/MESOS-1127


Repository: mesos


Description
---

Having a slave id will help us do better reconciliation. So added it to the 
'Reconcile' call.

Also updated the master to receive the Reconcile call directly.


Diffs (updated)
-

  include/mesos/scheduler/scheduler.proto 
783a63ad1cc0edd86605d638046fb959cb6e97e8 
  src/master/master.hpp c10e7c08c191acef9d31d98018a47a2a952a4dfc 
  src/master/master.cpp b1093bb867eb8624740f3d094cef2c23405a18b6 
  src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
  src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 

Diff: https://reviews.apache.org/r/32502/diff/


Testing
---

make check


Thanks,

Vinod Kone

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Alexander Rukletsov]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review80985
---

Ship it!



src/master/master.cpp


Is one space missing here?


- Alexander Rukletsov


On April 20, 2015, 7:59 p.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated April 20, 2015, 7:59 p.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp c10e7c08c191acef9d31d98018a47a2a952a4dfc 
>   src/master/master.cpp e30b951eda2b3b0d5b2a80716f0b32c6bbe041bc 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Ben Mahler]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review80827
---

Ship it!



include/mesos/scheduler/scheduler.proto


This still refers to 'statuses'?



src/master/master.cpp


How about some newlines here? You could set the slave id last since it 
requires an 'if'.



src/tests/scheduler_tests.cpp


Maybe we should test without this set?


- Ben Mahler


On April 20, 2015, 7:59 p.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated April 20, 2015, 7:59 p.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp c10e7c08c191acef9d31d98018a47a2a952a4dfc 
>   src/master/master.cpp e30b951eda2b3b0d5b2a80716f0b32c6bbe041bc 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/
---

(Updated April 20, 2015, 7:59 p.m.)


Review request for mesos and Ben Mahler.


Changes
---

Made SlaveID optional in RECONCILE.


Bugs: MESOS-1127
https://issues.apache.org/jira/browse/MESOS-1127


Repository: mesos


Description
---

Having a slave id will help us do better reconciliation. So added it to the 
'Reconcile' call.

Also updated the master to receive the Reconcile call directly.


Diffs (updated)
-

  include/mesos/scheduler/scheduler.proto 
783a63ad1cc0edd86605d638046fb959cb6e97e8 
  src/master/master.hpp c10e7c08c191acef9d31d98018a47a2a952a4dfc 
  src/master/master.cpp e30b951eda2b3b0d5b2a80716f0b32c6bbe041bc 
  src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
  src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 

Diff: https://reviews.apache.org/r/32502/diff/


Testing
---

make check


Thanks,

Vinod Kone

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

> On April 8, 2015, 5:52 p.m., Alexander Rukletsov wrote:
> > src/master/master.cpp, lines 3480-3482
> > 
> >
> > Do you want to update metrics here?
> > `++metrics->messages_reconcile_tasks;`

I'll add metrics in one sweep for all calls.


> On April 8, 2015, 5:52 p.m., Alexander Rukletsov wrote:
> > include/mesos/scheduler/scheduler.proto, line 191
> > 
> >
> > I see your point in making `SlaveID` required in `TaskStatus` and it 
> > makes sense to me. But does this mean it should be required here as well?
> > 
> > My understanding is that if the master knows `TaskID` it can always 
> > deduce `SlaveID`; if the task is unknown to the master, then the task is 
> > lost in all cases except when there are transitionary slaves. Having 
> > `SlaveID` in reconciliation query *just* speeds up reconciliation for 
> > unknown tasks, therefore making it required feels too much for me. Maybe 
> > Ben Mahler can provide more background and validate my thoughts.
> > 
> > If you opt for making it required, I would then suggest to add checks 
> > in reconciliation algorithm, so that `SlaveID` of known (to the master) 
> > tasks matches `SlaveID` from the query. Similar to what you do in a 
> > subsequent patch for `killTask`.
> > 
> > I'll also leave a comment on MESOS-1127.

After some discussion with benm an benh, i've decided to make SlaveID optional 
in both reconcileTasks and killTask for being backwards compatible with 
frameworks that haven't persisted SlaveId with their tasks.


> On April 8, 2015, 5:52 p.m., Alexander Rukletsov wrote:
> > include/mesos/mesos.proto, line 768
> > 
> >
> > For consistency, mind adding a comment in `Slave::statusUpdate()` that 
> > we should ensure `SlaveID` is set? Or even log warning in case it's not, so 
> > we start preparing framework authors for the upcoming change now. Also, 
> > maybe we need a JIRA for this.

see below. i'll remove the TODO for now.


> On April 8, 2015, 5:52 p.m., Alexander Rukletsov wrote:
> > src/master/master.cpp, line 3476
> > 
> >
> > I would suggest we put the implementation after `Master::accept()` call 
> > for consistency.

i'm keeping them close to their continuations for ease in readability in cpp.


- Vinod


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review79358
---


On April 3, 2015, 11:36 p.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated April 3, 2015, 11:36 p.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 3a8e8bf303e0576c212951f6028af77e54d93537 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
>   src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Alexander Rukletsov]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review79358
---



include/mesos/mesos.proto


For consistency, mind adding a comment in `Slave::statusUpdate()` that we 
should ensure `SlaveID` is set? Or even log warning in case it's not, so we 
start preparing framework authors for the upcoming change now. Also, maybe we 
need a JIRA for this.



include/mesos/scheduler/scheduler.proto


I see your point in making `SlaveID` required in `TaskStatus` and it makes 
sense to me. But does this mean it should be required here as well?

My understanding is that if the master knows `TaskID` it can always deduce 
`SlaveID`; if the task is unknown to the master, then the task is lost in all 
cases except when there are transitionary slaves. Having `SlaveID` in 
reconciliation query *just* speeds up reconciliation for unknown tasks, 
therefore making it required feels too much for me. Maybe Ben Mahler can 
provide more background and validate my thoughts.

If you opt for making it required, I would then suggest to add checks in 
reconciliation algorithm, so that `SlaveID` of known (to the master) tasks 
matches `SlaveID` from the query. Similar to what you do in a subsequent patch 
for `killTask`.

I'll also leave a comment on MESOS-1127.



src/master/master.cpp


I would suggest we put the implementation after `Master::accept()` call for 
consistency.



src/master/master.cpp


Do you want to update metrics here?
`++metrics->messages_reconcile_tasks;`


- Alexander Rukletsov


On April 3, 2015, 11:36 p.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated April 3, 2015, 11:36 p.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 3a8e8bf303e0576c212951f6028af77e54d93537 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
>   src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/
---

(Updated April 3, 2015, 11:36 p.m.)


Review request for mesos and Ben Mahler.


Changes
---

benm's comments.


Bugs: MESOS-1127
https://issues.apache.org/jira/browse/MESOS-1127


Repository: mesos


Description
---

Having a slave id will help us do better reconciliation. So added it to the 
'Reconcile' call.

Also updated the master to receive the Reconcile call directly.


Diffs (updated)
-

  include/mesos/mesos.proto 3a8e8bf303e0576c212951f6028af77e54d93537 
  include/mesos/scheduler/scheduler.proto 
783a63ad1cc0edd86605d638046fb959cb6e97e8 
  src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
  src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
  src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
  src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 

Diff: https://reviews.apache.org/r/32502/diff/


Testing
---

make check


Thanks,

Vinod Kone

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > There is another nice aspect of requiring SlaveID in Reconcile and in Kill, 
> > we can make SlaveID in TaskStatus required. Are you planning to add a TODO 
> > for that? Will we be ever be able to make it required?

Added a TODO. The only way we can make it required though would be to give 
advance notice to frameworks (executors) to always set it and then make it 
required in a future version.


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, lines 241-244
> > 
> >
> > Shouldn't this come before you create the mesos object?

REGISTERED event can only be received after a REGISTER call is sent. Just 
creating the Mesos object doesn't cause any registration to happen.


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, lines 241-244
> > 
> >
> > Shouldn't this come before you create the mesos object? Otherwise seems 
> > like you may receive the REGISTERED event before you set up this 
> > expectation?

see above.


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, line 309
> > 
> >
> > Do you want to check the reason?

good idea. done.


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, line 279
> > 
> >
> > newline here?

done


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, line 281
> > 
> >
> > newline here?

done


> On March 31, 2015, 9:49 p.m., Ben Mahler wrote:
> > src/tests/scheduler_tests.cpp, line 298
> > 
> >
> > newline here?

done


- Vinod


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review78391
---


On March 31, 2015, 12:07 a.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated March 31, 2015, 12:07 a.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
>   src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Ben Mahler]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review78391
---


There is another nice aspect of requiring SlaveID in Reconcile and in Kill, we 
can make SlaveID in TaskStatus required. Are you planning to add a TODO for 
that? Will we be ever be able to make it required?


include/mesos/scheduler/scheduler.proto


You can kill this now that we have a doc.



include/mesos/scheduler/scheduler.proto


Should we make this a Reconcile::Task as opposed to a Reconcile::Status, 
now that it's no longer a TaskStatus?

I'm also tempted to add a TODO for calling this 'Query' instead and 
allowing schedulers to query the master for additional things (other than 
tasks).



src/tests/scheduler_tests.cpp


Shouldn't this come before you create the mesos object?



src/tests/scheduler_tests.cpp


Shouldn't this come before you create the mesos object? Otherwise seems 
like you may receive the REGISTERED event before you set up this expectation?



src/tests/scheduler_tests.cpp


newline here?



src/tests/scheduler_tests.cpp


newline here?



src/tests/scheduler_tests.cpp


newline here?



src/tests/scheduler_tests.cpp


Do you want to check the reason?


- Ben Mahler


On March 31, 2015, 12:07 a.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated March 31, 2015, 12:07 a.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation. So added it to the 
> 'Reconcile' call.
> 
> Also updated the master to receive the Reconcile call directly.
> 
> 
> Diffs
> -
> 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
>   src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/
---

(Updated March 31, 2015, 12:07 a.m.)


Review request for mesos and Ben Mahler.


Changes
---

Updated description.


Bugs: MESOS-1127
https://issues.apache.org/jira/browse/MESOS-1127


Repository: mesos


Description (updated)
---

Having a slave id will help us do better reconciliation. So added it to the 
'Reconcile' call.

Also updated the master to receive the Reconcile call directly.


Diffs (updated)
-

  include/mesos/scheduler/scheduler.proto 
783a63ad1cc0edd86605d638046fb959cb6e97e8 
  src/master/master.hpp 05be911734b8d70c9fa5f3b4a275e8b610621fc5 
  src/master/master.cpp 618db68ee4163b06e479cf3413eda4b63c9c5a4b 
  src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
  src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 

Diff: https://reviews.apache.org/r/32502/diff/


Testing
---

make check


Thanks,

Vinod Kone

Re: Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Alexander Rojas]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/#review78035
---


I have to say, given the description and summary of the patch and its contents, 
I was surprised to see the code did much more than just adding the slave id, it 
also adds the logic to deal with the reconciliation case in the receive call. 
Can you please update the Description mentioning that?

Other than that, it looks good.

- Alexander Rojas


On March 26, 2015, 12:08 a.m., Vinod Kone wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/32502/
> ---
> 
> (Updated March 26, 2015, 12:08 a.m.)
> 
> 
> Review request for mesos and Ben Mahler.
> 
> 
> Bugs: MESOS-1127
> https://issues.apache.org/jira/browse/MESOS-1127
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Having a slave id will help us do better reconciliation.
> 
> 
> Diffs
> -
> 
>   include/mesos/scheduler/scheduler.proto 
> 783a63ad1cc0edd86605d638046fb959cb6e97e8 
>   src/master/master.hpp 3c957abcb54a0c23b8549c1d21d2d9277791938d 
>   src/master/master.cpp dccd7c635da4b7031cd109bd84e7f17b31777ef1 
>   src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
>   src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 
> 
> Diff: https://reviews.apache.org/r/32502/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Review Request 32502: Updated RECONCILE call to always specifiy slave id.

[Vinod Kone]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/32502/
---

Review request for mesos and Ben Mahler.


Bugs: MESOS-1127
https://issues.apache.org/jira/browse/MESOS-1127


Repository: mesos


Description
---

Having a slave id will help us do better reconciliation.


Diffs
-

  include/mesos/scheduler/scheduler.proto 
783a63ad1cc0edd86605d638046fb959cb6e97e8 
  src/master/master.hpp 3c957abcb54a0c23b8549c1d21d2d9277791938d 
  src/master/master.cpp dccd7c635da4b7031cd109bd84e7f17b31777ef1 
  src/scheduler/scheduler.cpp 584b042e32865fdf875bf41ebcfb7f9c327d882a 
  src/tests/scheduler_tests.cpp 4a89a7a88b50bb8c254f5076661ce07ac9fc7657 

Diff: https://reviews.apache.org/r/32502/diff/


Testing
---

make check


Thanks,

Vinod Kone