Re: Basebackup

Physical replication using PgBackrest would probably be even faster,
because parallelism will better use the network bandwidth.

On Mon, Sep 9, 2024 at 3:18 PM vrms <vrms(at)netcologne(dot)de> wrote:

> hi Rajesh,
>
> On 9/4/24 11:44 AM, Rajesh Kumar wrote:
>
> ... I have two openshift environments in different data centers. I want to
> move one data from one DC to another Data center.
>
> it is not 100% clear what exactly you want to achieve.
> But ... if you want to migrate your postgres from location A to location B
> (that is what it somehow sounds like) I guess ...
>
> 1. setting up a replica on the target (B) via pg_basebackup
> 2. stopping your application and the old main (A)
> 3. promoting the replica (B) to become the new main
> 4. re-directing your application to the target (B)
>
> could be quite efficient.
>
>
> On Wed, 4 Sept 2024, 13:49 Asad Ali, <asadalinagri(at)gmail(dot)com> wrote:
>
>> Hi Rajesh,
>>
>> Here are a few differences between basebackup and rsync.
>>
>> *pg_basebackup* is easier, as it's a single command with PostgreSQL
>> support. Data consistency is automatically ensured by postgresql. It
>> transfers the entire database, potentially high network bandwidth is
>> required. It is less flexible, due to mostly predefined options.
>>
>> *rsync *is more complex, it requires multiple manual steps. This method
>> requires stopping or ensuring the primary server's data directory is in a
>> consistent state before starting the synchronization, Which can cause
>> downtime or performance impact. It requires careful handling to ensure data
>> consistency. It is more efficient with bandwidth, as it only transfers
>> changes. It is more flexible with control over what gets copied.
>>
>> *You can use basebackup* if you want a straightforward, reliable way to
>> create a base backup. If you are looking for a method that integrates
>> directly with PostgreSQL's replication capabilities.
>> Network bandwidth and performance impact on the primary server are not
>> critical concerns. If you want an automated process that requires less
>> manual intervention.
>>
>> *You can use rsync* if you have specific requirements for how the data
>> should be copied (e.g., selective file copying). If you need more control
>> over the synchronization process or want to optimize for minimal network
>> usage. If you have a large dataset and want to minimize the data transfer
>> by only copying changes after an initial sync. If you are comfortable with
>> handling data consistency and can manage the additional complexity.
>>
>> Regards,
>> Asad Ali
>>
>>
>> On Tue, Sep 3, 2024 at 5:31 PM Rajesh Kumar <rajeshkumar(dot)dba09(at)gmail(dot)com>
>> wrote:
>>
>>> How different is basebackup from new server with primary host and rsync
>>> from primary, move to new server and start cluster?
>>>
>>

--
Death to America, and butter sauce.
Iraq lobster!