Re: Review Request 71194: Add `disk/xfs` isolator support for ephemeral volumes.

[Jiang Yan Xu]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71194/#review217058
---




src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 342 (patched)


Use `at()`?



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 346 (patched)


There are three experessions `foreachpair` so if we break them apart, put 
each one a line?



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 347 (patched)


s/container volumes/persistent volumes/?



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 423 (patched)


Do we need to single out the overlay backend? Seems like we can just scan 
all backends?



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 424 (patched)


FWIW sandbox scanning predates the container run state, now it does look 
like at least sandboxes (or any paths that can be deduced from the container 
ID) would be covered given it's checkpointed pretty early: 
https://github.com/apache/mesos/blob/e1176c453d04a8ef8f53cf23928b5bbb09173d78/src/slave/containerizer/mesos/containerizer.cpp#L1547

but of course the emphemeral volumes are added later than that and there 
could be dirs with projectIDs set before `ephemeral_volumes` is persisted.



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 517 (patched)


Same question: why don't we fail here?


- Jiang Yan Xu


On Aug. 1, 2019, 7:27 p.m., James Peach wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71194/
> ---
> 
> (Updated Aug. 1, 2019, 7:27 p.m.)
> 
> 
> Review request for mesos, Xudong Ni, Gilbert Song, Jie Yu, and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-9900
> https://issues.apache.org/jira/browse/MESOS-9900
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Add support for labeling ephemeral volumes with the sandbox XFS
> project ID. This makes changes to the container rootfs share the
> same disk quota as the sandbox.
> 
> 
> Diffs
> -
> 
>   src/slave/containerizer/mesos/isolators/xfs/disk.cpp 
> 646330c65b24aa28801ec99d7909db08a3e05c79 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.hpp 
> 362e02172d2fd8e6e241fb6f5689f569ba74a0d1 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.cpp 
> f2040cf36c601a13281a78ff844ebd41000a2d65 
> 
> 
> Diff: https://reviews.apache.org/r/71194/diff/3/
> 
> 
> Testing
> ---
> 
> sudo make check (Frdora 30)
> 
> 
> Thanks,
> 
> James Peach
> 
>

Re: Review Request 71192: Propagate ephemeral volume information from rootfs.

[Jiang Yan Xu]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71192/#review217057
---




src/slave/containerizer/mesos/provisioner/backend.hpp
Lines 55 (patched)


The backend abstraction probably shouldn't care about quota so probably 
phrasing it as "additional paths the backend has created" or something?



src/slave/containerizer/mesos/provisioner/backends/overlay.cpp
Lines 243 (patched)


Nit: with initializers I don't think we wrap elements with spaces.


- Jiang Yan Xu


On July 31, 2019, 6:03 p.m., James Peach wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71192/
> ---
> 
> (Updated July 31, 2019, 6:03 p.m.)
> 
> 
> Review request for mesos, Xudong Ni, Gilbert Song, Jie Yu, and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-9900
> https://issues.apache.org/jira/browse/MESOS-9900
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Propagate the overlayfs upperdir path to the containerization
> layer in a general form as a set of ephemeral volume paths.
> 
> 
> Diffs
> -
> 
>   include/mesos/slave/containerizer.proto 
> a60c96302a6cec90ecd0a0885b844fff8d37db71 
>   src/slave/containerizer/mesos/containerizer.cpp 
> a01edc8793a2eaa655f1729a01a01f1f61fbf7cb 
>   src/slave/containerizer/mesos/provisioner/backend.hpp 
> 7257d3a962ecdf87fe9d52facbd6a2619311a018 
>   src/slave/containerizer/mesos/provisioner/backends/aufs.hpp 
> 2c2518775d2bcb3b1857775723828c4ac08111ca 
>   src/slave/containerizer/mesos/provisioner/backends/aufs.cpp 
> 2eba55228e7596c170dddea2a9930ba6cf73404f 
>   src/slave/containerizer/mesos/provisioner/backends/bind.hpp 
> c4a1d5f35710f6eab13938333ed4cb97d9ff2f6b 
>   src/slave/containerizer/mesos/provisioner/backends/bind.cpp 
> 4cc4c520aa6872c05f474bc7837a353b43983e0b 
>   src/slave/containerizer/mesos/provisioner/backends/copy.hpp 
> 5dc9a2fd38cdfafda02129b0d144eee3230a1ff2 
>   src/slave/containerizer/mesos/provisioner/backends/copy.cpp 
> 10516caab82d910944814b80e0a1d9aeba19007c 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.hpp 
> 362e02172d2fd8e6e241fb6f5689f569ba74a0d1 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.cpp 
> f2040cf36c601a13281a78ff844ebd41000a2d65 
>   src/slave/containerizer/mesos/provisioner/provisioner.hpp 
> 7f84aa499b21140eb29ef7f81e2608743b12eb75 
>   src/slave/containerizer/mesos/provisioner/provisioner.cpp 
> a081fb059f9e0788c2153d1e04eda37fe514a1f7 
> 
> 
> Diff: https://reviews.apache.org/r/71192/diff/3/
> 
> 
> Testing
> ---
> 
> sudo make check (Frdora 30)
> 
> 
> Thanks,
> 
> James Peach
> 
>

Re: Review Request 71194: Add `disk/xfs` isolator support for ephemeral volumes.

[Jiang Yan Xu]

> On Aug. 2, 2019, 4:03 p.m., Jiang Yan Xu wrote:
> > src/slave/containerizer/mesos/isolators/xfs/disk.cpp
> > Lines 424 (patched)
> > 
> >
> > FWIW sandbox scanning predates the container run state, now it does 
> > look like at least sandboxes (or any paths that can be deduced from the 
> > container ID) would be covered given it's checkpointed pretty early: 
> > https://github.com/apache/mesos/blob/e1176c453d04a8ef8f53cf23928b5bbb09173d78/src/slave/containerizer/mesos/containerizer.cpp#L1547
> > 
> > but of course the emphemeral volumes are added later than that and 
> > there could be dirs with projectIDs set before `ephemeral_volumes` is 
> > persisted.

Sorry I meant to add that the provisioner dir for a container can also be 
[deduced](https://github.com/apache/mesos/blob/e1176c453d04a8ef8f53cf23928b5bbb09173d78/src/slave/containerizer/mesos/provisioner/paths.hpp#L38)
 from the containerID so you can just scan the subdirs of the containers 
recovered. 

Up to you.


- Jiang Yan


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71194/#review217058
---


On Aug. 1, 2019, 7:27 p.m., James Peach wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71194/
> ---
> 
> (Updated Aug. 1, 2019, 7:27 p.m.)
> 
> 
> Review request for mesos, Xudong Ni, Gilbert Song, Jie Yu, and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-9900
> https://issues.apache.org/jira/browse/MESOS-9900
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Add support for labeling ephemeral volumes with the sandbox XFS
> project ID. This makes changes to the container rootfs share the
> same disk quota as the sandbox.
> 
> 
> Diffs
> -
> 
>   src/slave/containerizer/mesos/isolators/xfs/disk.cpp 
> 646330c65b24aa28801ec99d7909db08a3e05c79 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.hpp 
> 362e02172d2fd8e6e241fb6f5689f569ba74a0d1 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.cpp 
> f2040cf36c601a13281a78ff844ebd41000a2d65 
> 
> 
> Diff: https://reviews.apache.org/r/71194/diff/3/
> 
> 
> Testing
> ---
> 
> sudo make check (Frdora 30)
> 
> 
> Thanks,
> 
> James Peach
> 
>

Re: Review Request 71193: Supported multiple quota paths in the `disk/xfs` isolator.

[Jiang Yan Xu]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71193/#review217059
---


Fix it, then Ship it!





src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Line 570 (original), 566 (patched)


You are not changing the logic here but could you remind me why this error 
doesn't fail the update?



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 776 (patched)


For `hashmap` there's already a `contains` defined.



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Lines 778-779 (patched)


You could've used a `else if` followed by an `else`  to avoid another level 
of nesting.



src/slave/containerizer/mesos/isolators/xfs/disk.cpp
Line 807 (original)


We could move the logging about the `dir` up above `erase` if we still want 
to log it (maybe for `VLOG(1)`)?


- Jiang Yan Xu


On July 30, 2019, 12:52 a.m., James Peach wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71193/
> ---
> 
> (Updated July 30, 2019, 12:52 a.m.)
> 
> 
> Review request for mesos, Xudong Ni, Gilbert Song, Jie Yu, and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-9900
> https://issues.apache.org/jira/browse/MESOS-9900
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> The `disk/xfs` isolator assumed that there would only be a single
> directory path for each project quota. When we apply project quotas
> to the overlayfs upperdir, that won't be true any more, since the
> upperdir will come under the same quota as the task sandbox.
> 
> Update the quota reclamation tracking to keep a set of disk paths that
> the quota has been applied to, and only reclaim the project ID once all
> those paths have been removed.
> 
> 
> Diffs
> -
> 
>   src/slave/containerizer/mesos/isolators/xfs/disk.hpp 
> 94d44e7f39dc01037461015b499a1fc3169b24e8 
>   src/slave/containerizer/mesos/isolators/xfs/disk.cpp 
> 646330c65b24aa28801ec99d7909db08a3e05c79 
> 
> 
> Diff: https://reviews.apache.org/r/71193/diff/3/
> 
> 
> Testing
> ---
> 
> sudo make check (Frdora 30)
> 
> 
> Thanks,
> 
> James Peach
> 
>

Re: Review Request 71192: Propagate ephemeral volume information from rootfs.

[Jiang Yan Xu]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71192/#review217060
---


Ship it!




Ship It!

- Jiang Yan Xu


On July 31, 2019, 6:03 p.m., James Peach wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71192/
> ---
> 
> (Updated July 31, 2019, 6:03 p.m.)
> 
> 
> Review request for mesos, Xudong Ni, Gilbert Song, Jie Yu, and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-9900
> https://issues.apache.org/jira/browse/MESOS-9900
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Propagate the overlayfs upperdir path to the containerization
> layer in a general form as a set of ephemeral volume paths.
> 
> 
> Diffs
> -
> 
>   include/mesos/slave/containerizer.proto 
> a60c96302a6cec90ecd0a0885b844fff8d37db71 
>   src/slave/containerizer/mesos/containerizer.cpp 
> a01edc8793a2eaa655f1729a01a01f1f61fbf7cb 
>   src/slave/containerizer/mesos/provisioner/backend.hpp 
> 7257d3a962ecdf87fe9d52facbd6a2619311a018 
>   src/slave/containerizer/mesos/provisioner/backends/aufs.hpp 
> 2c2518775d2bcb3b1857775723828c4ac08111ca 
>   src/slave/containerizer/mesos/provisioner/backends/aufs.cpp 
> 2eba55228e7596c170dddea2a9930ba6cf73404f 
>   src/slave/containerizer/mesos/provisioner/backends/bind.hpp 
> c4a1d5f35710f6eab13938333ed4cb97d9ff2f6b 
>   src/slave/containerizer/mesos/provisioner/backends/bind.cpp 
> 4cc4c520aa6872c05f474bc7837a353b43983e0b 
>   src/slave/containerizer/mesos/provisioner/backends/copy.hpp 
> 5dc9a2fd38cdfafda02129b0d144eee3230a1ff2 
>   src/slave/containerizer/mesos/provisioner/backends/copy.cpp 
> 10516caab82d910944814b80e0a1d9aeba19007c 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.hpp 
> 362e02172d2fd8e6e241fb6f5689f569ba74a0d1 
>   src/slave/containerizer/mesos/provisioner/backends/overlay.cpp 
> f2040cf36c601a13281a78ff844ebd41000a2d65 
>   src/slave/containerizer/mesos/provisioner/provisioner.hpp 
> 7f84aa499b21140eb29ef7f81e2608743b12eb75 
>   src/slave/containerizer/mesos/provisioner/provisioner.cpp 
> a081fb059f9e0788c2153d1e04eda37fe514a1f7 
> 
> 
> Diff: https://reviews.apache.org/r/71192/diff/3/
> 
> 
> Testing
> ---
> 
> sudo make check (Frdora 30)
> 
> 
> Thanks,
> 
> James Peach
> 
>

Review Request 71232: Renamed cmake parameter for parallel test execution.

[Benjamin Bannier]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71232/
---

Review request for mesos, Benno Evers and Till Toenshoff.


Bugs: MESOS-6382
https://issues.apache.org/jira/browse/MESOS-6382


Repository: mesos


Description
---

The Jenkins setup (which uses `support/docker-build.sh` under the
covers) is parameterized with the  `CONFIGURATION` environment variable.
While in we pass configure-style flags for both autotools and cmake
builds to it in the Jenkins
configuration, the script performs transformations so that
configure-style flags are transformed to cmake-style (replace `_` with
`-`, uppercase flags, replace `--` with `-D`).

We disable parallel test execution in Jenkins by passing
`--disable-parallel-test-execution` which with the transformations in
`support/docker-build.sh` leads to a cmake arg
`-DDISABLE_PARALLEL_TEST_EXECUTION=1`. This patch renames the cmake arg
from a default enabled `ENABLE_PARALLEL_TEST_EXECUTION` to a default
disabled `DISABLE_PARALLEL_TEST_EXECUTION` to support this workflow.


Diffs
-

  cmake/MesosConfigure.cmake ca8d406e1be9d6ab50c6c9dc4d2de97189b17b9a 


Diff: https://reviews.apache.org/r/71232/diff/1/


Testing
---

Tested ctest behavior with both the new flag enabled and disabled.


Thanks,

Benjamin Bannier

Re: Review Request 71150: Factored out storage provider method to update resources.

[Chun-Hung Hsiao]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71150/#review216947
---



I'm not sure if this is


src/resource_provider/storage/provider.cpp
Lines 292 (patched)


Did you accidentally add a new line?



src/resource_provider/storage/provider.cpp
Lines 728 (patched)


How about merging this function and `reconcileStoragePools` since the logic 
is very similar?



src/resource_provider/storage/provider.cpp
Lines 730 (patched)


Remove this debug log.



src/resource_provider/storage/provider.cpp
Lines 735 (patched)


Remove this debug log.


- Chun-Hung Hsiao


On July 23, 2019, 8:18 p.m., Benjamin Bannier wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71150/
> ---
> 
> (Updated July 23, 2019, 8:18 p.m.)
> 
> 
> Review request for mesos and Chun-Hung Hsiao.
> 
> 
> Bugs: MESOS-9254
> https://issues.apache.org/jira/browse/MESOS-9254
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Factored out storage provider method to update resources.
> 
> 
> Diffs
> -
> 
>   src/resource_provider/storage/provider.cpp 
> 6d632606f411d3ca99d3573a57c9f68b02ba8072 
> 
> 
> Diff: https://reviews.apache.org/r/71150/diff/1/
> 
> 
> Testing
> ---
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>

Re: Review Request 71232: Renamed cmake parameter for parallel test execution.

[Till Toenshoff via Review Board]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71232/#review217050
---


Fix it, then Ship it!





cmake/MesosConfigure.cmake
Line 49 (original), 49 (patched)


s/no/not/


- Till Toenshoff


On Aug. 2, 2019, 7:15 a.m., Benjamin Bannier wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71232/
> ---
> 
> (Updated Aug. 2, 2019, 7:15 a.m.)
> 
> 
> Review request for mesos, Benno Evers and Till Toenshoff.
> 
> 
> Bugs: MESOS-6382
> https://issues.apache.org/jira/browse/MESOS-6382
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> The Jenkins setup (which uses `support/docker-build.sh` under the
> covers) is parameterized with the  `CONFIGURATION` environment variable.
> While in we pass configure-style flags for both autotools and cmake
> builds to it in the Jenkins
> configuration, the script performs transformations so that
> configure-style flags are transformed to cmake-style (replace `_` with
> `-`, uppercase flags, replace `--` with `-D`).
> 
> We disable parallel test execution in Jenkins by passing
> `--disable-parallel-test-execution` which with the transformations in
> `support/docker-build.sh` leads to a cmake arg
> `-DDISABLE_PARALLEL_TEST_EXECUTION=1`. This patch renames the cmake arg
> from a default enabled `ENABLE_PARALLEL_TEST_EXECUTION` to a default
> disabled `DISABLE_PARALLEL_TEST_EXECUTION` to support this workflow.
> 
> 
> Diffs
> -
> 
>   cmake/MesosConfigure.cmake ca8d406e1be9d6ab50c6c9dc4d2de97189b17b9a 
> 
> 
> Diff: https://reviews.apache.org/r/71232/diff/1/
> 
> 
> Testing
> ---
> 
> Tested ctest behavior with both the new flag enabled and disabled.
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>

Re: Review Request 71151: Performed periodic storage local provider reconciliations.

[Chun-Hung Hsiao]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71151/#review217049
---




src/resource_provider/storage/provider.cpp
Lines 749-768 (patched)


```
loop(
  self(),
  std::bind(::after, reconciliationInterval),
  [this](const Nothing&) {
// Poll resource provider state ...
reconciled = sequence.add(defer(self(), ::reconcileResources));

return reconciled.then([](const Nothing&) -> ControlFlow {
  return Continue();
});
  });
```



src/resource_provider/storage/provider.cpp
Lines 756-760 (original), 801-821 (patched)


How about introduce a `bool alwaysUpdate` parameter in this function:
```
bool shouldUpdate = alwaysUpdate;

if (result != totalResources) {
  LOG(INFO) << "Removing"... ;
  
  // Update the resource version...
  totalResources = result;
  resourceVersion = id::UUID::random();
  checkpointResourceProviderState();
  
  shouldUpdate = true;
}

if (shouldUpdate) {
  sendResourceProviderStateUpdate();
}
```

Then, in `reconcileResourceProviderState()`:
```
  return reconcileOperationStatuses()
.then(defer(self(), ::reconcileResources, true))
.then(defer(self(), [this] {
  statusUpdateManager.resume();
  
  LOG(INFO)
<< "Resource provider " << info.id() << " is in READY state";

  state = READY;

  return Nothing();
}));
```
So we don't put the state transition logic inside this function?



src/resource_provider/storage/provider.cpp
Lines 1356 (patched)


The delay won't guarantee the happens-before order between the first 
reconciliation and the ones in `watchResources`. How about:

```
reconciled = reconcileResourceProviderState()
  .onReady(defer(self(), ::watchProfiles))
  .onReady(defer(self(), ::watchResources))
  .onFailed(...)
  .onDiscarded(...);



src/resource_provider/storage/provider.cpp
Lines 1869-1886 (original), 1933-1950 (patched)


With periodic reconciliation, this won't be necessary. Removing this would 
also reduce a source of reconciliation, so the reconciliation can only happen 
in either `watchProfiles` or `watchResources`.


- Chun-Hung Hsiao


On July 29, 2019, 8:56 a.m., Benjamin Bannier wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71151/
> ---
> 
> (Updated July 29, 2019, 8:56 a.m.)
> 
> 
> Review request for mesos and Chun-Hung Hsiao.
> 
> 
> Bugs: MESOS-9254
> https://issues.apache.org/jira/browse/MESOS-9254
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> Performed periodic storage local provider reconciliations.
> 
> 
> Diffs
> -
> 
>   src/resource_provider/storage/provider.cpp 
> 6d632606f411d3ca99d3573a57c9f68b02ba8072 
>   src/tests/storage_local_resource_provider_tests.cpp 
> 69b59d48ceefebbb7accefe411c54ac5cecff1c3 
> 
> 
> Diff: https://reviews.apache.org/r/71151/diff/2/
> 
> 
> Testing
> ---
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>

Re: Review Request 71232: Renamed cmake parameter for parallel test execution.

[Mesos Reviewbot]

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71232/#review217051
---



Patch looks great!

Reviews applied: [71232]

Passed command: export OS='ubuntu:14.04' BUILDTOOL='autotools' COMPILER='gcc' 
CONFIGURATION='--verbose --disable-libtool-wrappers 
--disable-parallel-test-execution' ENVIRONMENT='GLOG_v=1 MESOS_VERBOSE=1'; 
./support/docker-build.sh

- Mesos Reviewbot


On Aug. 2, 2019, 12:15 a.m., Benjamin Bannier wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71232/
> ---
> 
> (Updated Aug. 2, 2019, 12:15 a.m.)
> 
> 
> Review request for mesos, Benno Evers and Till Toenshoff.
> 
> 
> Bugs: MESOS-6382
> https://issues.apache.org/jira/browse/MESOS-6382
> 
> 
> Repository: mesos
> 
> 
> Description
> ---
> 
> The Jenkins setup (which uses `support/docker-build.sh` under the
> covers) is parameterized with the  `CONFIGURATION` environment variable.
> While in we pass configure-style flags for both autotools and cmake
> builds to it in the Jenkins
> configuration, the script performs transformations so that
> configure-style flags are transformed to cmake-style (replace `_` with
> `-`, uppercase flags, replace `--` with `-D`).
> 
> We disable parallel test execution in Jenkins by passing
> `--disable-parallel-test-execution` which with the transformations in
> `support/docker-build.sh` leads to a cmake arg
> `-DDISABLE_PARALLEL_TEST_EXECUTION=1`. This patch renames the cmake arg
> from a default enabled `ENABLE_PARALLEL_TEST_EXECUTION` to a default
> disabled `DISABLE_PARALLEL_TEST_EXECUTION` to support this workflow.
> 
> 
> Diffs
> -
> 
>   cmake/MesosConfigure.cmake ca8d406e1be9d6ab50c6c9dc4d2de97189b17b9a 
> 
> 
> Diff: https://reviews.apache.org/r/71232/diff/2/
> 
> 
> Testing
> ---
> 
> Tested ctest behavior with both the new flag enabled and disabled.
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>