Re: why not using a mountpoint as PGDATA?

On 2/27/19 12:43 PM, Joe Conway wrote:
> On 2/27/19 11:49 AM, Peter J. Holzer wrote:
>> On 2019-02-27 10:42:12 -0500, Tom Lane wrote:
>>> Luca Ferrari <fluca1978(at)gmail(dot)com> writes:
>>>> On Wed, Feb 27, 2019 at 12:33 PM Julien Rouhaud <rjuju123(at)gmail(dot)com> wrote:
>>>>> You can see most obvious reasons at
>>>>> https://bugzilla.redhat.com/show_bug.cgi?id=1247477
>> [...]
>>> The case that I can recall most clearly was actually in the other
>>> direction: during system bootup, some NFS volume that was being abused
>>> this way (mount point == data dir) was slow to mount. Compounding the
>>> problem, postgres was being started through some init script that would
>>> helpfully run initdb if it saw the specified data directory was empty.
>>> So, rather than failing like a properly paranoid DBA would wish, it
>>> ran initdb and then started the postmaster.
>> Ouch.
>>
>> I wonder though why that directory was writable by the postgres user.
>> But maybe the helpful start script chown'ed it to fix the "wrong"
>> permissions.
> FWIW, if you want to read the whole gory details of that incident, here
> it is:
>
> https://www.postgresql.org/message-id/flat/41D04FA4.7010402%40joeconway.com#dfc38927745e238d49569ffd5b33beba

What in the world was that SuSE maintainer -- and the people who should have
been looking over his shoulder -- thinking???

--
Angular momentum makes the world go 'round.