[Bucardo-announce] entire bucardo breaks if 1 dbgroup mbr offline

[Vilem Kebrt]

Hi Jonathan , can you send us error information from log ?
I think i made some fix about this to version 4.3 few years ago, seems 
this could be similar problem.
We were using this type of configuration for Radius postgresql db and if 
i'm correct people i was that time working with are still using it. 
Maybe it can be problem of some changes in perl from that times, we were 
using FreeBSD on db servers and there are no such major changes.
Vilem

On 11.10.2013 16:40, Jonathan Brinkman wrote:
>
> Greetings all.
>
> I am still looking for a solution to this. Using Ubuntu 12.04 with 
> Bucardo 4.5.0
>
> I created a push-delta sync to a dbgroup consisting of 2 members. If 
> either of those members goes offline and bucardo is restarted, it 
> breaks not just that push-delta sync but ALL syncs, including those 
> not related to the offline member. I think its b/c it never validates 
> the broken sync, and the restart process dies every time.
>
> How can I keep everyone else up while the bad boy stays offline? This 
> makes the setup NOT scalable. I was hoping to use it with 30+ dbgroup 
> members, but not with that kind of fragility.
>
> What other options exist? Does bucardo 5 fix this, and I should bite 
> the bullet and upgrade?
>
>
>
> _______________________________________________
> Bucardo-announce mailing list
> Bucardo-announce at bucardo.org
> https://mail.endcrypt.com/mailman/listinfo/bucardo-announce

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.endcrypt.com/pipermail/bucardo-announce/attachments/20131011/2d183f19/attachment.html>