This is really nice for SoHo segment, till a time when some will use Piwigo with 60k photos and delivered DB will stop performing because needs more performance from diff host (what is a common situation). Because with Piwigo I can choose my own db instance.
We are a kubernetes native project, where we aim that the
complete solution is rolled out in the same kubernetes cluster.
As we explained earlier we view the first release of SCALE (22.02) as SOHO only, as it does not support multi-node kubernetes clusters. For the later release of SCALE supporting kubernetes clusters, we will be working on both HA databases and spreading of load.
In our opinion, it's not yet ready for SMB/SME use.
but this "single monolithic App" idea is in contradiction with your explanation:
It's not a monolithic App. The databases are seperate containers in the backend, that can be loadbalanced accordingly, which, as explained, will be more thorougly employed on future SCALE releases.
It's also not in contradiction with anything, because such features are always nicely hidden behind advanced settings checkboxes.
We offer a LOT of options most users will never use. That doesn't mean they are worthless, just because you won't use them.
We are also already working our butts off, getting closer to SMB/SME grade. Just because a future is useless for most people now, doesn't mean it doesn't have a place in our design for the future.
for Unifi controller running on Mongo DB, there is more suitable use ui.community directly
No idea what you are trying to say here.
where is an added value for the SME/SMB environment
We don't offer an SME/SMB product yet, as explained a few times now.
Nor does iX Systems when it comes to SCALE Apps, in our opinion.
just the only possible way to use containers modified for k3s from you?
We, generally, don't build the containers (yet). There are some future drafts for going (more) in that direction, but those are not anywhere close to being actively worked on.
How you will cover the situation when some would like to use Telegraf+InfluxDB+Grafana but with Influx version 1.8 and not in 2.0 because the new one uses diff. Flux language which isn't compatible?
In case a breaking change is just breaking because a company restructured stuff, we aim to make migration as easy as possible.
In cases like Influx 1 to Influxdb 2, it would be considered a completely new product (As that's how other projects view it as well), hence there would just be InfluxDB-1 and InfluxDB-2.
Question:
- it means, that all databases inside the "monolithic" containers use the same setup of root, psw, db names, ... ?
We at TrueCharts (as can be read in one of our latest posts thoroughly laying out our security views), take security serieusly.
Obviously passwords are randomised. DB names and usernames are not randomised however.
In case a user wants additional security, the best way of doing that would be implementing networkPolicies in the future.
---
All being said:
We are our own community, with our own goals and idea's. To which we are (somewhat) accountable.
So we will leave it at this. If you've any input on our project, you can converse with our community directly, using the usual channels.
K.S.