Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 24, 2015, 6:15 p.m., Jie Yu wrote:
> > FYI, there's a regression in this patch and I committed a fix:
> > 
> > commit 46f80f270bc335172e919343f6f14e007216823f
> > Author: Jie Yu 
> > Date:   Tue Mar 24 11:06:08 2015 -0700
> > 
> > Fixed a regression caused by the slave checkpoint flag removal.
> > 
> > diff --git a/src/tests/port_mapping_tests.cpp 
> > b/src/tests/port_mapping_tests.cpp
> > index 8192dea..623840e 100644
> > --- a/src/tests/port_mapping_tests.cpp
> > +++ b/src/tests/port_mapping_tests.cpp
> > @@ -1669,7 +1669,6 @@ public:
> >ContainerizerTest::CreateSlaveFlags();
> >  
> >  // Setup recovery slave flags.
> > -flags.checkpoint = true;
> >  flags.recover = "reconnect";
> >  flags.strict = true;
> 
> Adam B wrote:
> Thanks for finding this Jie. I believe flags.recover and flags.strict are 
> the defaults, so we can simplify this function to just
> `return ContainerizerTest::CreateSlaveFlags();`
> Joerg, would you like to handle this cleanup?

Sure, see https://issues.apache.org/jira/browse/MESOS-2538.


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77608
---


On March 19, 2015, 3:52 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 3:52 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 24, 2015, 11:15 a.m., Jie Yu wrote:
> > FYI, there's a regression in this patch and I committed a fix:
> > 
> > commit 46f80f270bc335172e919343f6f14e007216823f
> > Author: Jie Yu 
> > Date:   Tue Mar 24 11:06:08 2015 -0700
> > 
> > Fixed a regression caused by the slave checkpoint flag removal.
> > 
> > diff --git a/src/tests/port_mapping_tests.cpp 
> > b/src/tests/port_mapping_tests.cpp
> > index 8192dea..623840e 100644
> > --- a/src/tests/port_mapping_tests.cpp
> > +++ b/src/tests/port_mapping_tests.cpp
> > @@ -1669,7 +1669,6 @@ public:
> >ContainerizerTest::CreateSlaveFlags();
> >  
> >  // Setup recovery slave flags.
> > -flags.checkpoint = true;
> >  flags.recover = "reconnect";
> >  flags.strict = true;

Thanks for finding this Jie. I believe flags.recover and flags.strict are the 
defaults, so we can simplify this function to just
`return ContainerizerTest::CreateSlaveFlags();`
Joerg, would you like to handle this cleanup?


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77608
---


On March 19, 2015, 8:52 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 8:52 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Jie Yu]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77608
---


FYI, there's a regression in this patch and I committed a fix:

commit 46f80f270bc335172e919343f6f14e007216823f
Author: Jie Yu 
Date:   Tue Mar 24 11:06:08 2015 -0700

Fixed a regression caused by the slave checkpoint flag removal.

diff --git a/src/tests/port_mapping_tests.cpp b/src/tests/port_mapping_tests.cpp
index 8192dea..623840e 100644
--- a/src/tests/port_mapping_tests.cpp
+++ b/src/tests/port_mapping_tests.cpp
@@ -1669,7 +1669,6 @@ public:
   ContainerizerTest::CreateSlaveFlags();
 
 // Setup recovery slave flags.
-flags.checkpoint = true;
 flags.recover = "reconnect";
 flags.strict = true;

- Jie Yu


On March 19, 2015, 3:52 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 3:52 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 19, 2015, 9:33 a.m., Adam B wrote:
> > Looks great! I'll try to get this committed later today, if nobody else has 
> > any objections.

Committed. @joerg84 please mark this review as "Submitted"


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77055
---


On March 19, 2015, 8:52 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 8:52 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 19, 2015, 4:33 p.m., Adam B wrote:
> > src/tests/fault_tolerance_tests.cpp, line 123
> > 
> >
> > No need to call it a "checkpointing slave", since all slaves are 
> > checkpointing now.
> 
> Joerg Schad wrote:
> Till was initially wondering why the behavior changed here, so I wanted 
> to emphasize... but I can remove it...
> 
> Adam B wrote:
> No big deal. Your call.

Then I would leave it, as it makes it a little easier to understand the change.


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77055
---


On March 19, 2015, 3:52 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 3:52 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 19, 2015, 9:33 a.m., Adam B wrote:
> > src/tests/fault_tolerance_tests.cpp, line 123
> > 
> >
> > No need to call it a "checkpointing slave", since all slaves are 
> > checkpointing now.
> 
> Joerg Schad wrote:
> Till was initially wondering why the behavior changed here, so I wanted 
> to emphasize... but I can remove it...

No big deal. Your call.


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77055
---


On March 19, 2015, 8:52 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 8:52 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 19, 2015, 4:33 p.m., Adam B wrote:
> > src/tests/fault_tolerance_tests.cpp, line 123
> > 
> >
> > No need to call it a "checkpointing slave", since all slaves are 
> > checkpointing now.

Till was initially wondering why the behavior changed here, so I wanted to 
emphasize... but I can remove it...


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77055
---


On March 19, 2015, 3:52 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 3:52 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77055
---

Ship it!


Looks great! I'll try to get this committed later today, if nobody else has any 
objections.


src/tests/fault_tolerance_tests.cpp


No need to call it a "checkpointing slave", since all slaves are 
checkpointing now.


- Adam B


On March 19, 2015, 8:52 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 8:52 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77054
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 19, 2015, 3:52 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 19, 2015, 3:52 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 19, 2015, 3:52 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Adressed Adam's comment and style review.


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs (updated)
-

  include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
  src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
  src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX (had 
to exclude some known flaky tests under OSX)


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 19, 2015, 6:33 a.m., Adam B wrote:
> > src/tests/master_allocator_tests.cpp, line 707
> > 
> >
> > Should this line also get a comment explaining why shutdown=true?

Done (+ fixed style issue)


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77022
---


On March 18, 2015, 9:43 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 18, 2015, 9:43 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77022
---

Ship it!


LGTM! Just some comments about comments, and then I'd be happy to commit it.


src/tests/status_update_manager_tests.cpp


s/when checking the checkpoint/when recovering the checkpointed data/



src/tests/fault_tolerance_tests.cpp


Something sounds weird with "...as otherwise with checkpointing..."
How about "Stop the slave with an explicit shutdown message so that the 
master sees it exit cleanly."



src/tests/master_allocator_tests.cpp


Should this line also get a comment explaining why shutdown=true?



src/tests/master_authorization_tests.cpp


.. with an explicit shutdown message...



src/tests/master_tests.cpp


How about "// Reuse slaveFlags so next StartSlave() uses the same 
work_dir."?



src/tests/master_tests.cpp


Style: Comments wrap at 70 chars.



src/tests/master_tests.cpp


Is this needed for its work_dir?


- Adam B


On March 18, 2015, 2:43 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 18, 2015, 2:43 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 18, 2015, 11:04 p.m., Adam B wrote:
> >

(Sorry, pulled the trigger early; meant to click "Save" instead. More coming..)


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77020
---


On March 18, 2015, 2:43 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 18, 2015, 2:43 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review77020
---



src/tests/master_authorization_tests.cpp


Nit: s/> >/>>/ or is that coming in a separate review?


- Adam B


On March 18, 2015, 2:43 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 18, 2015, 2:43 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76967
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 18, 2015, 9:43 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 18, 2015, 9:43 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
>   src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
>   src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 18, 2015, 9:43 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Adressed Adam's comment regarding CreateSlaveFlags (removed where unessary and 
added comments otherwise)


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs (updated)
-

  include/mesos/mesos.proto ec8efaec13f54a56d82411f6cdbdb8ad8b103748 
  src/slave/flags.hpp dbaf5f532d0bc65a6d16856b8ffcc2c06a98f1fa 
  src/slave/slave.cpp f1f210045e6100560f0d26244f9675f4543a5620 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX (had 
to exclude some known flaky tests under OSX)


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 15, 2015, 9:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 2018
> > 
> >
> > Do we even need to CreateSlaveFlags() here and elsewhere? If you're not 
> > setting any non-default flag values or otherwise using the 'slaveFlags' 
> > variable, it can be removed, since StartSlave() is the same as 
> > StartSlave(CreateSlaveFlags()).
> 
> Joerg Schad wrote:
> CreateSlaveFlags also generates a new work_dir assignment, as I restart 
> the slave I woud like to keep the same work_dir.
> 
> Adam B wrote:
> Sure, but StartSlave(None()) still calls 
> cluster.slaves.start(CreateSlaveFlags()), so it's implicit.
> ```
> Try > MesosTest::StartSlave(
> const Option& flags)
> {
>   return cluster.slaves.start(
>   flags.isNone() ? CreateSlaveFlags() : flags.get());
> }
> ```

Just when we restart the slave a new work_dir will be created as there will be 
a second implicit call to createSlaveFlags. I will add a comment.


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76502
---


On March 16, 2015, 10:07 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 16, 2015, 10:07 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 15, 2015, 2:03 a.m., Adam B wrote:
> > include/mesos/mesos.proto, lines 321-323
> > 
> >
> > Will we be able to remove this flag in 0.23, or will we need to wait 
> > for another release cycle for deprecation? Seems like if it was already 
> > 'optional' in 0.22, and it was never set/saved as true in 0.22, then we 
> > could remove it in 0.23, as an 0.23 slave recovering 0.22 SlaveInfo would 
> > be fine. What about an 0.22 slave registering with an 0.23 master, and vice 
> > versa? How safe would that be?
> > Maybe we need to change the default to true here?
> 
> Joerg Schad wrote:
> Is it ok if I move this discussion to the follow-up Jira as it is not 
> really an issue with this patch?

Sure. Definitely not a blocker for this patch, but I would like to see it 
covered by another JIRA.


> On March 15, 2015, 2:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 1928
> > 
> >
> > Not yours, but could you help out our style update push and s/> >/>>/ 
> > in code next to your changes (not necessarily the entire file)?
> 
> Joerg Schad wrote:
> As seemingly there is some discussion whether to do such style fixes or 
> not, I will provide another Patch/Jira to fix > > -> >> for all tests.
> 
> Till Toenshoff wrote:
> I would suggest to not do that. Let me try to explain:
> 
> This particual RR is rather unusual in that it touches many files. In 
> october of last year, we reached a consesus for style debt fixes; we said 
> that it would be nice if all files touched would also get this style update 
> (dev-list: `Large changes on the codebase due to MESOS-1872`).
> 
> A. Lets not bundle any style debt fixes with this one at all as it is 
> atypical and not covered by our consesus.
> 
> B. Lets make two RRs out of it, first fixing all sharp bracket whitspaces 
> on the files to be touched by this RR. Then base this RR on those style 
> fixes. This will still allow quick and easy reviews but also get a big pile 
> of style debt fixes merged into the project.

SGTM.


> On March 15, 2015, 2:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 2018
> > 
> >
> > Do we even need to CreateSlaveFlags() here and elsewhere? If you're not 
> > setting any non-default flag values or otherwise using the 'slaveFlags' 
> > variable, it can be removed, since StartSlave() is the same as 
> > StartSlave(CreateSlaveFlags()).
> 
> Joerg Schad wrote:
> CreateSlaveFlags also generates a new work_dir assignment, as I restart 
> the slave I woud like to keep the same work_dir.

Sure, but StartSlave(None()) still calls 
cluster.slaves.start(CreateSlaveFlags()), so it's implicit.
```
Try > MesosTest::StartSlave(
const Option& flags)
{
  return cluster.slaves.start(
  flags.isNone() ? CreateSlaveFlags() : flags.get());
}
```


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76502
---


On March 16, 2015, 3:07 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 16, 2015, 3:07 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_test

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Till Toenshoff]

> On March 15, 2015, 9:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 1928
> > 
> >
> > Not yours, but could you help out our style update push and s/> >/>>/ 
> > in code next to your changes (not necessarily the entire file)?
> 
> Joerg Schad wrote:
> As seemingly there is some discussion whether to do such style fixes or 
> not, I will provide another Patch/Jira to fix > > -> >> for all tests.

I would suggest to not do that. Let me try to explain:

This particual RR is rather unusual in that it touches many files. In october 
of last year, we reached a consesus for style debt fixes; we said that it would 
be nice if all files touched would also get this style update (dev-list: `Large 
changes on the codebase due to MESOS-1872`).

A. Lets not bundle any style debt fixes with this one at all as it is atypical 
and not covered by our consesus.

B. Lets make two RRs out of it, first fixing all sharp bracket whitspaces on 
the files to be touched by this RR. Then base this RR on those style fixes. 
This will still allow quick and easy reviews but also get a big pile of style 
debt fixes merged into the project.


- Till


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76502
---


On March 16, 2015, 10:07 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 16, 2015, 10:07 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

> On March 15, 2015, 9:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 2018
> > 
> >
> > Do we even need to CreateSlaveFlags() here and elsewhere? If you're not 
> > setting any non-default flag values or otherwise using the 'slaveFlags' 
> > variable, it can be removed, since StartSlave() is the same as 
> > StartSlave(CreateSlaveFlags()).

CreateSlaveFlags also generates a new work_dir assignment, as I restart the 
slave I woud like to keep the same work_dir.


> On March 15, 2015, 9:03 a.m., Adam B wrote:
> > include/mesos/mesos.proto, lines 321-323
> > 
> >
> > Will we be able to remove this flag in 0.23, or will we need to wait 
> > for another release cycle for deprecation? Seems like if it was already 
> > 'optional' in 0.22, and it was never set/saved as true in 0.22, then we 
> > could remove it in 0.23, as an 0.23 slave recovering 0.22 SlaveInfo would 
> > be fine. What about an 0.22 slave registering with an 0.23 master, and vice 
> > versa? How safe would that be?
> > Maybe we need to change the default to true here?

Is it ok if I move this discussion to the follow-up Jira as it is not really an 
issue with this patch?


> On March 15, 2015, 9:03 a.m., Adam B wrote:
> > src/tests/master_tests.cpp, line 1928
> > 
> >
> > Not yours, but could you help out our style update push and s/> >/>>/ 
> > in code next to your changes (not necessarily the entire file)?

As seemingly there is some discussion whether to do such style fixes or not, I 
will provide another Patch/Jira to fix > > -> >> for all tests.


- Joerg


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76502
---


On March 16, 2015, 10:07 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 16, 2015, 10:07 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76542
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 16, 2015, 10:07 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 16, 2015, 10:07 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 16, 2015, 10:07 a.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Adressed Adam's comments


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs (updated)
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX (had 
to exclude some known flaky tests under OSX)


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

> On March 2, 2015, 11:39 a.m., Cody Maloney wrote:
> > src/tests/master_allocator_tests.cpp, line 637
> > 
> >
> > The test shouldn't be deleted, rather updated to make sure that after 
> > the timeout of a slave being really considered lost, the resources are 
> > removed.
> > 
> > Side note: It would probably be good for tasks to have a "must report 
> > back time" seperate from when we consider a slave truly lost. For quick one 
> > off-jobs, waiting minutes before retrying the task will be unacceptable 
> > latency... Probably some sort of early "slave disappeared" message to 
> > frameworks when we first detect, then a framework can re-launch if needed...
> 
> Joerg Schad wrote:
> see comment for above dropped test.

Cody: If the Master actually receives an "Exited" event for this pid, it will 
mark it as disconnected/deactivated and not offer any resources from it, but 
during a network disconnect there is no way to know if the slave has exited or 
if the task is still running.


- Adam


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review74795
---


On March 14, 2015, 7:10 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 14, 2015, 7:10 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Adam B]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76502
---


First pass looks good. I've got a couple of questions about the follow-up 
removal of SlaveInfo.checkpoint, and then some style nits and general cleanup 
requests. I still need to go through the actual test logic itself to understand 
those changes, but the core change looks good.


include/mesos/mesos.proto


Style nit: s/remove/Remove/



include/mesos/mesos.proto


Will we be able to remove this flag in 0.23, or will we need to wait for 
another release cycle for deprecation? Seems like if it was already 'optional' 
in 0.22, and it was never set/saved as true in 0.22, then we could remove it in 
0.23, as an 0.23 slave recovering 0.22 SlaveInfo would be fine. What about an 
0.22 slave registering with an 0.23 master, and vice versa? How safe would that 
be?
Maybe we need to change the default to true here?



include/mesos/mesos.proto


s/Mesos/MESOS/



src/slave/slave.cpp


s/as to be/to be/



src/slave/slave.cpp


Minor: Technically, "Delete after 0.23.", since you cannot complete the 
deprecation during 0.23.1 or another point release.



src/tests/docker_containerizer_tests.cpp


All of these lines can be removed like you did above, right?



src/tests/gc_tests.cpp


s/  / / (remove double-space), and you can also s/the garbage/garbage/



src/tests/master_authorization_tests.cpp


s/  / / (remove double-space)



src/tests/master_tests.cpp


s/FLags/flags/



src/tests/master_tests.cpp


s/> >/>>/



src/tests/master_tests.cpp


Remove the now-unecessary comment, and these default-value initializations 
(as you did elsewhere).



src/tests/master_tests.cpp


Not yours, but could you help out our style update push and s/> >/>>/ in 
code next to your changes (not necessarily the entire file)?



src/tests/master_tests.cpp


Do we even need to CreateSlaveFlags() here and elsewhere? If you're not 
setting any non-default flag values or otherwise using the 'slaveFlags' 
variable, it can be removed, since StartSlave() is the same as 
StartSlave(CreateSlaveFlags()).



src/tests/partition_tests.cpp


Remove the CreateSlaveFlags() line if nothing else uses 'flags'.
s/> >/>>/



src/tests/slave_recovery_tests.cpp


Should be able to remove all of these too, right? Then you can just 
directly `return ContainerizerTest::CreateSlaveFlags();`


- Adam B


On March 14, 2015, 7:10 a.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 14, 2015, 7:10 a.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f853

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review76488
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 14, 2015, 2:10 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 14, 2015, 2:10 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX 
> (had to exclude some known flaky tests under OSX)
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 14, 2015, 2:10 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Adressed Till's (unpublished) comments.


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs (updated)
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 0f99e4efb8fa2b96f120a3e49191158ca0364c06 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp e69348be676a80017062e3abbd15b8008a6009d7 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX (had 
to exclude some known flaky tests under OSX)


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 7, 2015, 12:32 a.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

more Testing


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing (updated)
---

make check GTEST_BREAK_ON_FAILURE=1 GTEST_SHUFFLE=1 GTEST_REPEAT=50 on OSX (had 
to exclude some known flaky tests under OSX)


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review75580
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 6, 2015, 10:39 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 6, 2015, 10:39 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 6, 2015, 10:39 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Changed comment to adress cody's comment.


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs (updated)
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Cody Maloney]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review75540
---


LGTM!

Need some mesos committers to review now.


src/slave/slave.cpp


"Checkpointing of slaves is always enabled."


- Cody Maloney


On March 6, 2015, 2:22 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 6, 2015, 2:22 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Mesos ReviewBot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/#review75488
---


Patch looks great!

Reviews applied: [31539]

All tests passed.

- Mesos ReviewBot


On March 6, 2015, 2:22 p.m., Joerg Schad wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31539/
> ---
> 
> (Updated March 6, 2015, 2:22 p.m.)
> 
> 
> Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.
> 
> 
> Bugs: MESOS-2375
> https://issues.apache.org/jira/browse/MESOS-2375
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> As a number of tests rely on the checkpointing flag to be false, a few tests 
> had to be adapted.
> Removed the following test as the tested logic is specific to (old) 
> non-checkpointing slaves:
> SlaveRecoveryTest.NonCheckpointingSlave:
>This test checks whether a non-checkpointing slave is not scheduled to a 
> checkpointing framework.   
>It can be removed as all slaves are checkpointing slaves.
> 
> 
> Diffs
> -
> 
>   include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
>   src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
>   src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
>   src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
>   src/tests/docker_containerizer_tests.cpp 
> 06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
>   src/tests/fault_tolerance_tests.cpp 
> 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
>   src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
>   src/tests/master_allocator_tests.cpp 
> a432d0207e1a92532a495bf9ad2826414ee4f6f0 
>   src/tests/master_authorization_tests.cpp 
> ff706ed6f8537207b30a548b0ce2121c5df71ab9 
>   src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
>   src/tests/master_validation_tests.cpp 
> c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
>   src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
>   src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
>   src/tests/persistent_volume_tests.cpp 
> b617117ade4b487cc06002cfeca76a0486833b20 
>   src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
>   src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
>   src/tests/status_update_manager_tests.cpp 
> 216a22e9f292b4141c8b966dad0f25dbd791c025 
> 
> Diff: https://reviews.apache.org/r/31539/diff/
> 
> 
> Testing
> ---
> 
> make check
> 
> 
> Thanks,
> 
> Joerg Schad
> 
>

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 6, 2015, 2:22 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

updated description


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description (updated)
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
SlaveRecoveryTest.NonCheckpointingSlave:
   This test checks whether a non-checkpointing slave is not scheduled to a 
checkpointing framework.   
   It can be removed as all slaves are checkpointing slaves.


Diffs
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 6, 2015, 12:48 p.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description (updated)
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following test as the tested logic is specific to (old) 
non-checkpointing slaves:
-SlaveRecoveryTest.NonCheckpointingSlave
Checks whether a non-checkpointing slave is not scheduled to a checkpointing 
framework.
All slaves are checkpointing slaves.


Diffs
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check


Thanks,

Joerg Schad

Re: Review Request 31539: Remove the checkpoint variable entirely from slave/flags.hpp.

[Joerg Schad]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31539/
---

(Updated March 6, 2015, 11:38 a.m.)


Review request for mesos, Adam B, Cody Maloney, and Till Toenshoff.


Changes
---

Adressed Cody's comments.


Bugs: MESOS-2375
https://issues.apache.org/jira/browse/MESOS-2375


Repository: mesos


Description
---

As a number of tests rely on the checkpointing flag to be false, a few tests 
had to be adapted.
Removed the following three tests as the tested logic is specific to (old) 
non-checkpointing slaves:
-SlaveRecoveryTest.NonCheckpointingSlave
Checks whether a non-checkpointing slave is not scheduled to a checkpointing 
framework.
All slaves are checkpointing slaves.
-GarbageCollectorIntegrationTest.Restart
Checks whether GC removes workings directories after a slave is lost.
With checkpointing we want to keep the checkpointing data.
-MasterAllocatorTest.SlaveLost
Checks whether resources of killed slave are recovered and not offered again.
Resources are held for a certain period of time to allow checkpoint recovery.


Diffs (updated)
-

  include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8 
  src/slave/flags.hpp 56b25caf3901b38bdecb50310e8bcae0b114efa8 
  src/slave/slave.cpp 2d52ea0c440fd530174b0e44a59c8ae68fa2616c 
  src/tests/disk_quota_tests.cpp 9c3a8815c3478535b72888c296a4aa5cda341ba3 
  src/tests/docker_containerizer_tests.cpp 
06cd3d89ecbaaac17ae6970604b21fbe29f6e887 
  src/tests/fault_tolerance_tests.cpp 9ac75b1f601e14a3d3d117775f37a4a48b291dc6 
  src/tests/gc_tests.cpp deaa6b1b6c32ae6d153229248d7d4f57caa0ebcf 
  src/tests/master_allocator_tests.cpp a432d0207e1a92532a495bf9ad2826414ee4f6f0 
  src/tests/master_authorization_tests.cpp 
ff706ed6f8537207b30a548b0ce2121c5df71ab9 
  src/tests/master_tests.cpp 580e1f818201f951c11e4e652a7941fcd888356d 
  src/tests/master_validation_tests.cpp 
c8742928a4e93e86ccd0f5a39856a65cfe8eb74f 
  src/tests/mesos.cpp c8f43d21b214e75eaac2870cbdf4f03fd18707d1 
  src/tests/partition_tests.cpp bb96aed37861867fbde68445016f0c6e039f3fb4 
  src/tests/persistent_volume_tests.cpp 
b617117ade4b487cc06002cfeca76a0486833b20 
  src/tests/reconciliation_tests.cpp acd70021574b05ab23872add5bdfa4a46b7dfc51 
  src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1 
  src/tests/status_update_manager_tests.cpp 
216a22e9f292b4141c8b966dad0f25dbd791c025 

Diff: https://reviews.apache.org/r/31539/diff/


Testing
---

make check


Thanks,

Joerg Schad