Hi Sylvain,
Thanks for the reply.
I would like to use the droplet implementation. This was the first thing I
looked at, however we are running Ubuntu 18.04 exclusively and the packages are
not available for this distro.
Are there any plans from bareos to create these packages for Ubuntu?
Met
Hi Rick,
Sorry for the delay of my answer, but I just discover your question.
We were using same configuration, but we have had a lot of troubles with it
(problems of striper, repetitive crashes, slowness, ...).
We have reconfigured BAREOS for an S3 approach on our CEPH cluster.
So, you have
Probably. That's why I'd probably prefer some duplicated, but not
necessarily HA setup. I'm always a bit suspicious towards HA-clustering
non-HA-capable solutions.
But YMMV
On 05.06.2020 14:47, Oleg Volkov wrote:
Jobs will be obviously aborted and failed.
Then you have to care about them
Hi,
it would be highly appreciated if bareos could send notifications via
webhooks or web API. AFAIK "Messages" directive only supports email.
Can "Mail Command" be customized to use eg cURL?
Let say.. something similar to:
Messages {
Name = webhook_ERM
Mail Command = "curl -vs -H
Jobs will be obviously aborted and failed.
Then you have to care about them manually as usual for failed job.
K.O.
On Friday, June 5, 2020 at 1:32:40 PM UTC+3, Spadajspadaj wrote:
>
> I would be, however, cautious about possible scenarios where a node breaks
> and fails over to the other
Thank you a lot, this helps me!
Kind regards,
Birgit
On 05/06/20 09:25, Spadajspadaj wrote:
If you don't specify retention period, they will get set at default
values so it's not a proper solution. I'd rather go and set it to some
insanely huge value.
But of course it will result in
I would be, however, cautious about possible scenarios where a node
breaks and fails over to the other server - for example - in the middle
of a backup job. Such scenarios would need some testing so you know what
to expect and how to handle such situation.
On 05.06.2020 09:00, Oleg Volkov
If you don't specify retention period, they will get set at default
values so it's not a proper solution. I'd rather go and set it to some
insanely huge value.
But of course it will result in ever-growing storage demand for the
catalog database since no jobs/files/volumes will be getting
I do not see any problem. It is just a service.
Make postgres HA, make /etc/bareos and /var/lib/bareos be on shared disk,
make VIP and colocate it with bareos services.
Never tried this with bareos, but made a lot of clusters - there should be
no problem.
Just follow any active-standby
