Re: BUG: Unable to handle kernel paging request for data at address __percpu_counter_add

[Raghavendra K T <raghavendra.kt@linux.vnet.ibm.com>]

On 11/24/2015 02:43 AM, Tejun Heo wrote:
> Hello,
>
> On Thu, Nov 19, 2015 at 03:54:35PM +0530, Raghavendra K T wrote:
>> While I was creating thousands of docker container on a power8 baremetal
>> (config: 4.3.0 kernel 1TB RAM, 20core (=160 cpu) system. After creating
>> around 5600 container
>> I have hit below problem.
>> [This is looking similar to
>> https://bugzilla.kernel.org/show_bug.cgi?id=101011, but
>> kernel had Revert "ext4: remove block_device_ejected" (bdfe0cbd746aa9) since
>> it is 4.3.0 tagged kernel]
>>
>> Any hints on how to go about the fix. Please let me know if you think any
>> more information needed.
>>
>> docker daemon is device mapper based. (and it took a day to recreate the
>> problem)
>>
>> [ by disabling  CONFIG_BLK_CGROUP and CONFIG_CGROUP_WRITEBACK I am able to
>> create 10k containers without any problem]
>
> Could be the same problem that Ilya is trying to fix.  ie. blkdev i_wb
> pointing to a stale wb.  Can you please see whether the following
> patch resolves the issue?
>
>   http://lkml.kernel.org/g/1448054554-24138-1-git-send-email-idryomov@gmail.com
>

Hi Tejun,

Thanks again for the pointer. I was now able to create more than 10k
containers without any problem  with CGROUP_WRITEBACK on whereas
earlier I had hit this problem few times around 5k+ containers itself.
(Also Replying to Ilya's thread).