Re: DSM robustness failure (was Re: Peripatus/failures)

Thomas Munro <thomas.munro@enterprisedb.com>

From: Thomas Munro <thomas.munro@enterprisedb.com>

To: Tom Lane <tgl@sss.pgh.pa.us>

Cc: Larry Rosenman <ler@lerctr.org>, Pg Hackers <pgsql-hackers@postgresql.org>, Robert Haas <robertmhaas@gmail.com>, Amit Kapila <amit.kapila16@gmail.com>

Date: 2018-10-17T22:08:33Z

Lists: pgsql-hackers

Attachments

0001-Fix-thinko-in-DSM-control-segment-crash-restart-logi.patch (application/octet-stream) patch 0001

On Thu, Oct 18, 2018 at 9:43 AM Thomas Munro
<thomas.munro@enterprisedb.com> wrote:
> On Thu, Oct 18, 2018 at 9:00 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> > I would argue that both dsm_postmaster_shutdown and dsm_postmaster_startup
> > are broken here; the former because it makes no attempt to unmap
> > the old control segment (which it oughta be able to do no matter how badly
> > broken the contents are), and the latter because it should not let
> > garbage old state prevent it from establishing a valid new segment.
>
> Looking.

(CCing Amit Kapila)

To reproduce this, I attached lldb to a backend and did "mem write
&dsm_control->magic 42", and then delivered SIGKILL to the backend.
Here's one way to fix it.  I think we have no choice but to leak the
referenced segments, but we can free the control segment.  See
comments in the attached patch for rationale.

-- 
Thomas Munro
http://www.enterprisedb.com