abudnik commented on a change in pull request #355: Handle EBUSY when
destroying a cgroup.
URL: https://github.com/apache/mesos/pull/355#discussion_r401684158
##########
File path: src/linux/cgroups.cpp
##########
@@ -696,13 +696,24 @@ Try<Nothing> remove(const string& hierarchy, const
string& cgroup)
string path = path::join(hierarchy, cgroup);
// Do NOT recursively remove cgroups.
- Try<Nothing> rmdir = os::rmdir(path, false);
- if (rmdir.isError()) {
- return Error(
+ // TODO The retry was added as a fix for kernel bug
+ // https://lkml.org/lkml/2020/1/15/1349
+ // where calling rmdir on a seemingly empty cgroup can fail
+ // with EBUSY while some tasks are exiting.
+ auto delay = Milliseconds(1);
+ for (auto retry = 10; ;) {
+ Try<Nothing> rmdir = os::rmdir(path, false);
+ if (!rmdir.isError()) {
+ return rmdir;
+ } else if (retry > 0) {
+ os::sleep(delay);
Review comment:
Note that `os::sleep()` is a blocking call while the `remove()` function is
supposed to be non-blocking. Calling a blocking function from a non-blocking
context must be avoided if possible.
If the error occurs for the `freezer` cgroup only,
[Destroyer::remove](https://github.com/apache/mesos/blob/f8a3dd334934094ec44e07fa350f958d218bc78f/src/linux/cgroups.cpp#L1536-L1555)
method can be modified to asynchronously retry "remove cgroup" operation.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
