[GENERAL] hot backup with zfs ?

2015-08-20 Thread Laurent Laborde
Friendly greetings !

I'm wondering if anyone tried the following hot backup process, and if it
works :
pg_start_backup()
zfs snapshop
pg_stop_backup()

copy the snapshot to the backup server.
Thx :)

PS : the WAL will still be archived the usual way.

-- 
Laurent "ker2x" Laborde


Re: [GENERAL] hot backup with zfs ?

2015-08-21 Thread Laurent Laborde
On Thu, Aug 20, 2015 at 4:57 PM, Joseph Kregloh 
wrote:

>
>
> On Thu, Aug 20, 2015 at 9:50 AM, Laurent Laborde 
> wrote:
>
>> Friendly greetings !
>>
>> I'm wondering if anyone tried the following hot backup process, and if it
>> works :
>> pg_start_backup()
>> zfs snapshop
>> pg_stop_backup()
>>
>> copy the snapshot to the backup server.
>>
>
Thx Joseph and Yves :)

-- 
Laurent "ker2x" Laborde


[GENERAL] pgcon2015, what happened to SMR disk technolgy ?

2017-10-17 Thread Laurent Laborde
Friendly greetings !

i remember an interesting talk from seagate at pgcon2015 about SMR disk
technology, and i use them for archive & backup (personal usage).

However, the highest capacity on the seagate archive product line (the one
using SMR) is 8TB.
Seagate have a 8TB ironwolf product at roughly the same price.
And a 12TB ironwolf, much more expensive since it's new, but 12TB
nonetheless.

But their SMR disk are still maxed at 8TB.

What's the point of the seagate archive now ?
Ironwolf, for the same public price, have better performance (obviously)
and, more surprising, a better MTBF.

I'm confused ...

Thank you :)


-- 
Laurent "ker2x" Laborde


Re: [GENERAL] pgcon2015, what happened to SMR disk technolgy ?

2017-10-17 Thread Laurent Laborde
On Tue, Oct 17, 2017 at 1:38 PM, Geoff Winkless  wrote:

> On 17 October 2017 at 11:59, Laurent Laborde  wrote:
>
>> What's the point of the seagate archive now ?
>> Ironwolf, for the same public price, have better performance (obviously)
>> and, more surprising, a better MTBF.
>>
>
> ​I have no real insight into whether Seagate are still pursuing the
> product design, but I'm not really surprised that the MTBF is worse: if the
> shingled disk must write some tracks twice for each individual track write,
> it seems logical that there will be more write stress and therefore
> shortened lifespan, no?
>

I contacted seagate and just got a reply : they don't have strategic
information to share about SMR technology at the moment.
I guess i saw it coming ^^


-- 
Laurent "ker2x" Laborde


Re: [GENERAL] Log storage

2017-10-18 Thread Laurent Laborde
Friendly greetings !

You may want to take a look at a postgresql "fork" called pipelinedb :
https://www.pipelinedb.com/
https://github.com/pipelinedb/pipelinedb

I'm not working for them, not using it, but i happen to know it exist :)

*hugs*

-- 
Laurent "ker2x" Laborde