Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8254854

[cgroups v1] Metric limits not properly detected on some join controller combinations

    Details

    • Subcomponent:
    • Resolved In Build:
      b22
    • CPU:
      generic
    • OS:
      linux

      Backports

        Description

        This bug is similar to JDK-8217766. In fact, it had an incomplete fix for the Metrics (Java) side.

        On a system with join controllers, /proc/self/cgroup, might look like this:

        9:pids:/user.slice/user-1000.slice/session-2.scope
        8:perf_event:/
        7:blkio:/user.slice
        6:rdma:/
        5:cpuset:/
        4:devices:/user.slice
        3:cpu,cpuacct,memory,net_cls,net_prio,hugetlb:/user.slice/user-1000.slice/session-2.scope
        2:freezer:/
        1:name=systemd:/user.slice/user-1000.slice/session-2.scope
        0::/user.slice/user-1000.slice/session-2.scope

        Then, the Java code to set the path to for the controller reads:

                try (Stream<String> lines =
                        CgroupUtil.readFilePrivileged(Paths.get("/proc/self/cgroup"))) {

                    lines.map(line -> line.split(":"))
                         .filter(line -> (line.length >= 3))
                         .forEach(line -> setSubSystemControllerPath(subsystem, line));

                } catch (IOException e) {
                    return null;
                }


        where setSubSystemController() reads:

        /**
             * setSubSystemPath based on the contents of /proc/self/cgroup
             */
            private static void setSubSystemControllerPath(CgroupV1Subsystem subsystem, String[] entry) {
                String controllerName;
                String base;
                CgroupV1SubsystemController controller = null;
                CgroupV1SubsystemController controller2 = null;

                controllerName = entry[1];
                base = entry[2];
                if (controllerName != null && base != null) {
                    switch (controllerName) {
                        case "memory":
                            controller = subsystem.memoryController();
                            break;
                        case "cpuset":
                            controller = subsystem.cpuSetController();
                            break;
                        case "cpu,cpuacct":
                        case "cpuacct,cpu":
                            controller = subsystem.cpuController();
                            controller2 = subsystem.cpuAcctController();
                            break;
                        case "cpuacct":
                            controller = subsystem.cpuAcctController();
                            break;
                        case "cpu":
                            controller = subsystem.cpuController();
                            break;
                        case "blkio":
                            controller = subsystem.blkIOController();
                            break;
                        // Ignore subsystems that we don't support
                        default:
                            break;
                    }
                }

                if (controller != null) {
                    controller.setPath(base);
                    if (controller instanceof CgroupV1MemorySubSystemController) {
                        CgroupV1MemorySubSystemController memorySubSystem = (CgroupV1MemorySubSystemController)controller;
                        boolean isHierarchial = getHierarchical(memorySubSystem);
                        memorySubSystem.setHierarchical(isHierarchial);
                        boolean isSwapEnabled = getSwapEnabled(memorySubSystem);
                        memorySubSystem.setSwapEnabled(isSwapEnabled);
                    }
                if (controller != null) {
                    controller.setPath(base);
                    if (controller instanceof CgroupV1MemorySubSystemController) {
                        CgroupV1MemorySubSystemController memorySubSystem = (CgroupV1MemorySubSystemController)controller;
                        boolean isHierarchial = getHierarchical(memorySubSystem);
                        memorySubSystem.setHierarchical(isHierarchial);
                        boolean isSwapEnabled = getSwapEnabled(memorySubSystem);
                        memorySubSystem.setSwapEnabled(isSwapEnabled);
                    }
                    subsystem.setActiveSubSystems();
                }
                if (controller2 != null) {
                    controller2.setPath(base);
                }

        So for the example /proc/self/cgroup file it only sets the path for "blkio" and "cpuset". Others are not correctly set because they are on a joined path.

        The net effect of this is that Metrics doesn't report the container limits correctly on such systems.

          Attachments

            Issue Links

              Activity

                People

                • Assignee:
                  sgehwolf Severin Gehwolf
                  Reporter:
                  sgehwolf Severin Gehwolf
                • Votes:
                  0 Vote for this issue
                  Watchers:
                  4 Start watching this issue

                  Dates

                  • Created:
                    Updated:
                    Resolved: