Checking for empty cgroup is not correct, because init process leaves cgroup early in do_exit. This leads to a situation, when container is treated as stopped but its resources (VEIP for instance) are not yet released. Which in turn leads to container restart failure due to non-releases VEIP address.
https://jira.sw.ru/browse/PSBM-78078 Signed-off-by: Stanislav Kinsburskiy <skinsbur...@virtuozzo.com> --- kernel/ve/ve.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/ve/ve.c b/kernel/ve/ve.c index b0188c3..c628516 100644 --- a/kernel/ve/ve.c +++ b/kernel/ve/ve.c @@ -846,7 +846,7 @@ static int ve_state_read(struct cgroup *cg, struct cftype *cft, if (ve->is_running) seq_puts(m, "RUNNING"); - else if (!nr_threads_ve(ve)) + else if (!ve->init_task) seq_puts(m, "STOPPED"); else if (ve->ve_ns) seq_puts(m, "STOPPING"); _______________________________________________ Devel mailing list Devel@openvz.org https://lists.openvz.org/mailman/listinfo/devel
