RHEL 8.8/8.6(EUS): hung_task_timeout_secs at migration_entry_wait_on_locked
Issue
- 升级到 RHEL 8.8 后,有个别命令进入挂起状态,系统平均负载显示非常高。
- 在回溯中,日志显示了几个带有
migration_entry_wait_on_locked()
的 hung_task_timeout_secs。
INFO: task task1:1618 blocked for more than 120 seconds.
Not tainted 4.18.0-477.10.1.el8_8.x86_64 #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:task1 state:D stack: 0 pid: 1618 ppid: 1 flags:0x00000080
Call Trace:
__schedule+0x2d1/0x870
schedule+0x55/0xf0
io_schedule+0x12/0x40
migration_entry_wait_on_locked+0x1ea/0x290
do_swap_page+0x5b0/0x710
__handle_mm_fault+0x453/0x6c0
handle_mm_fault+0xca/0x2a0
__do_page_fault+0x1f0/0x450
do_page_fault+0x37/0x130
page_fault+0x1e/0x30
- 升级到 RHEL 8.6 EUS
kernel-4.18.0-372.91.1.el8_6
后,个别命令进入挂起状态,系统平均负载显示非常高。 - 在回溯中,日志显示了几个带有
migration_entry_wait_on_locked()
的 hung_task_timeout_secs。
INFO: task task2:2159306 blocked for more than 120 seconds.
Tainted: GOE ----------- 4.18.0-372.91.1.el8_6.x86_64 #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:task2 state:D stack: 0 pid:2159306 ppid: 1 flags:0x00000084
Call Trace:
__schedule+0x2d1/0x860
schedule+0x55/0xf0
io_schedule+0x12/0x40
migration_entry_wait_on_locked+0x1e0/0x280
do_swap_page+0x5b0/0x710
__handle_mm_fault+0x453/0x6d0
handle_mm_fault+0xca/0x2a0
__do_page_fault+0x1d0/0x420
do_page_fault+0x37/0x12d
page_fault+0x1e/0x30
- 可能需要硬重置,因为问题可能阻止系统正常重启。
Environment
- Red Hat Enterprise Linux 8.8
kernel-4.18.0-477.10.1.el8_8
- Red Hat Enterprise Linux 8.6 EUS
kernel-4.18.0-372.91.1.el8_6
或更高版本
- Red Hat OpenShift Container Platform 4.12.49
kernel-4.18.0-372.89.1.el8_6
- Red Hat OpenShift Container Platform 4.12.50
kernel-4.18.0-372.89.1.el8_6
- Red Hat OpenShift Container Platform 4.12.51
kernel-4.18.0.372.93.1.el8_6
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.