On Wed, Feb 21, 2018 at 4:50 PM, Peter Geoghegan wrote:
> On Wed, Feb 21, 2018 at 7:53 AM, Rick Otten wrote:
>> side note: The disadvantage of local SSD is that it won't survive "hitting
>> the virtual power button" on an instance, nor can it migrate automatically
>> to other hardware. (We have
On Wed, Feb 21, 2018 at 12:07 PM, Claudio Freire wrote:
> On Wed, Feb 21, 2018 at 4:50 PM, Peter Geoghegan wrote:
>> On Wed, Feb 21, 2018 at 7:53 AM, Rick Otten wrote:
>>> side note: The disadvantage of local SSD is that it won't survive "hitting
>>> the virtual power button" on an instance, no
On Wed, Feb 21, 2018 at 7:53 AM, Rick Otten wrote:
> side note: The disadvantage of local SSD is that it won't survive "hitting
> the virtual power button" on an instance, nor can it migrate automatically
> to other hardware. (We have to hit the power button to add memory/cpu to
> the system, an
On Wed, Feb 21, 2018 at 7:53 AM, Rick Otten
wrote:
> Some of my data processes use large quantities of temp space - 5 or 6T
> anyway.
>
> We are running in Google Cloud. In order to get the best performance out
> of all of my queries that might need temp space, I've configured temp space
> on a
Rick Otten writes:
> At the moment it seems like when multiple temp spaces are available, the
> temp spaces are chosen in a 'round robin' or perhaps 'random' fashion. Is
> that true?
Yes, see fd.c's SetTempTablespaces and GetNextTempTableSpace.
There's no concept of different temp spaces having
Some of my data processes use large quantities of temp space - 5 or 6T
anyway.
We are running in Google Cloud. In order to get the best performance out
of all of my queries that might need temp space, I've configured temp space
on a concatenated local (volatile) SSD volume. In GCE, local SSD's a
