[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20180914161841.zqenjczqgfkycdmm@mac.bytemobile.com>
Date: Fri, 14 Sep 2018 18:18:41 +0200
From: Roger Pau Monné <roger.pau@...rix.com>
To: Valentin Vidic <Valentin.Vidic@...Net.hr>
CC: <drbd-user@...ts.linbit.com>, Jens Axboe <axboe@...nel.dk>,
Konrad Rzeszutek Wilk <konrad.wilk@...cle.com>,
<linux-kernel@...r.kernel.org>, <stable@...r.kernel.org>,
<linux-block@...r.kernel.org>, <xen-devel@...ts.xenproject.org>
Subject: Re: [DRBD-user] [PATCH] xen-blkback: Switch to closed state after
releasing the backing device
On Fri, Sep 14, 2018 at 01:49:51PM +0200, Valentin Vidic wrote:
> On Thu, Sep 13, 2018 at 05:08:00PM +0200, Roger Pau Monné wrote:
> > So I have the following patch which I think might solve your issues
> > while keeping the reset logic working. Would you mind giving it a try
> > with your use case and pvgrub?
>
> Thanks for the patch. It seems to be having some problems with pvgrub:
> machines don't boot past pvgrub and they are using 100% CPU. Also in
> dom0 xenstored is usind 100% CPU - strace reports it is looping in this:
>
> read(26, "\\\240\0\0", 4) = 4
> read(26, "\0\0\0\0L\0\0\0)\0\0\0\35\0\0\0+\\\232`\231\31\1&", 24) = 24
> read(26, "/local/domain/0/backend/vbd/1/51712/state", 41) = 41
> read(26, "\2\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\1\0\0\0\1\0\0\0004", 29) = 29
> read(26, "\0\330\0\0", 4) = 4
> read(26, "\0\0\0\0L\0\0\0)\0\0\0\35\0\0\0+\\JG\231\31\1&", 24) = 24
> read(26, "/local/domain/0/backend/vbd/2/51712/state", 41) = 41
> read(26, "\2\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\2\0\0\0\1\0\0\0004", 29) = 29
> read(26, "\0\200\1\0", 4) = 4
> read(26, "\0\0\0\0L\0\0\0)\0\0\0\35\0\0\0+\\Z\373\231\31\1&", 24) = 24
> read(26, "/local/domain/0/backend/vbd/5/51712/state", 41) = 41
> read(26, "\2\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\5\0\0\0\1\0\0\0004", 29) = 29
> read(12, "\321\1\0\0", 4) = 4
> write(12, "\321\1\0\0", 4) = 4
I see the issue, blkfront in pvgrub is waiting for the backend to
switch to Closed state before booting the OS kernel, so my fix is not
going to work. I guess I will have to find a proper way to switch to
Closed state.
Thanks, Roger.
Powered by blists - more mailing lists