Hello Guys,
I met a ocfs2 hang issue when running on Multi-nodes Refcount Block Group Test
on three nodes cluster, this hang issue can not reproduced each time, but it
looks there is a bug in dlmglue layer.
The bug symptom looks as below,
Node1:
ocfs2te+ 2403 2372 2403 01 Jan18 ?
Hi Gang,
Never met such a lockup issue so far with o2cb stack applied.
And I don't know how to debug this with fs/dlm.
What I suspect currently is that this lock resource was once QUEUED but fail to
re-QUEUED since BLOCKED has been marked.
Thanks,
Changwei
On 2018/1/24 14:30, Gang He wrote:
>
Hi Changwei,
>>>
> Hi Gang,
>
> Never met such a lockup issue so far with o2cb stack applied.
> And I don't know how to debug this with fs/dlm.
I think this issue should belong to dlmglue layer problem.
> What I suspect currently is that this lock resource was once QUEUED but fail
> to re-QUE