Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (1.0)
Subject: Re: OT: Performance of VM
From: Andrew Kerber <andrew.kerber@gmail.com>
In-Reply-To: 
 <CAM9pMnMQGTBPgbHSYE91NUNcZsrryTWbQ1ior0M-=Bcvk-CVKQ@mail.gmail.com>
Date: Sat, 10 Feb 2018 08:58:41 -0600
Cc: Andreas Kretschmer <andreas@a-kretschmer.de>,
 pgsql-performance@lists.postgresql.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <17D5011A-5D52-454F-B8A6-BE69CDD0878B@gmail.com>
References: <3bc54eb3-135d-692c-47e2-208189a350d7@thomas-guettler.de>
 <91fec1f8-d11d-2413-6b57-5310ab0e3b8a@a-kretschmer.de>
 <CAJvnOJbzmAiwytvpAkLaDC2Y4p=VS+1A_2qou7pyBU5bQPX5CQ@mail.gmail.com>
 <CAM9pMnMQGTBPgbHSYE91NUNcZsrryTWbQ1ior0M-=Bcvk-CVKQ@mail.gmail.com>
To: Robert Klemme <shortcutter@googlemail.com>
Precedence: bulk

I am consultant that specializes in virtualizing oracle enterprise level wor=
kloads.  I=E2=80=99m picking up Postgres as a secondary skill.  You are righ=
t if you don=E2=80=99t manage it properly, you can have problems running ent=
erprise workloads on vm s. But it can be done with proper management. And th=
e HA and DR advantages of virtual systems are huge.=20

Sent from my iPhone

> On Feb 10, 2018, at 5:20 AM, Robert Klemme <shortcutter@googlemail.com> wr=
ote:
>=20
>> On Mon, Feb 5, 2018 at 5:22 PM, Andrew Kerber <andrew.kerber@gmail.com> w=
rote:
>> Have them check the memory and CPU allocation of the hypervisor, make sur=
e
>> its not overallocated. Make sure the partitions for stroage are aligned (=
see
>> here:
>> https://blogs.vmware.com/vsphere/2011/08/guest-os-partition-alignment.htm=
l)
>> . Install tuned, and enable the throughput performance profile. Oracle ha=
s a
>> problem with transparent hugepages, postgres may well have the same probl=
em,
>> so consider disabling transparent hugepages.  There is no reason why
>> performance on a VM would be worse than performance on a physical server.=

>=20
> Not theoretically. But in practice if you have anything run in a VM
> like in this case you do not know what else is working on that box.
> Analyzing these issues can be really cumbersome and tricky. This is
> why I am generally skeptical of running a resource intensive
> application like a RDBMS in a VM. To get halfway predictable results
> you want at least a minimum of resources (CPU, memory, IO bandwidth)
> reserved for that VM.
>=20
> Anecdote: we once had a customer run our application in a VM (which is
> supported) and complain about slowness. Eventually we found out that
> they over committed memory - not in sum for all VMs which is common,
> but this single VM had been configured to have more memory than was
> physically available in the machine.
>=20
> Kind regards
>=20
> robert
>=20
> --=20
> [guy, jim, charlie].each {|him| remember.him do |as, often| as.you_can
> - without end}
> http://blog.rubybestpractices.com/