Postgres BDR EOL

So I am not the one making these decisions, my boss is and I am tasked with finding an alternative for our new cluster build. Is there a reason as to why it is okay to continue using BDR that I can report back with?
 
I underestimated my superiors willingness to accept this as an answer. That being said is there any replacement for BDR in the roadmap for Fusion? I recall Mark saying in one of his Advanced Classes videos that he was working on something to replace BDR, but I can find nothing at all about alternatives.
 

ad5ou

Active Member
The best answer to report back with "why keep using it" is because it will keep working. :p

Mark has mentioned several times he planned on finding an alternative replication method but has not shared any progress.

I haven't tried it yet but Bucardo might work as an alternative.
 

DigitalDaz

Administrator
Staff member
I don't understand why the 'EOL' tag means anything in this context. I have BDR running since 2015 and have never had to touch it. We have to use an old version of Postgres in order to use it anyway. Though I have never used it, I do see a number of voices raising the question of Galera.

With regards to BDR, apart from a typo in an error message that was fixed, BDR 1 hasn't been touched since April anyway.
 

markjcrane

Active Member
Staff member
I underestimated my superiors willingness to accept this as an answer. That being said is there any replacement for BDR in the roadmap for Fusion? I recall Mark saying in one of his Advanced Classes videos that he was working on something to replace BDR, but I can find nothing at all about alternatives.
Yes its definitely on the roadmap. We should be able to keep using BDR until then.