The feedback between developers and users seems to be broken.

[urza]

I am TrueNAS user for many years (started with NAS4Free, then FreeNAS, then TrueNAS Core and now I am trying to transition to TreuNAS Scale).

But I always felt like the devs are somewhat distant, not present in the community at all. Often times when I had some issue, it was already experienced by someone else before me, sometime many years before and the issue still existed.

I am experiencing the same feeling now as I am trying to transition to TrueNAS Scale.

I had two issues in recent week:
1. A problem with VM and TreuNAS Scale being able to see each other. There are multiple forum posts of people having the same issue dating back to 2020:

TrueNAS Scale can't connect to VM

I have a TrueNAS-SCALE-22.02.0.1 running with a VM Ubuntu Server. I can't ping my VM from the TrueNAS shell. root@truenas[~]# ping 192.168.1.26 PING 192.168.1.26 (192.168.1.26) 56(84) bytes of data. From 192.168.1.253 icmp_seq=1 Destination Host Unreachable But from my windows PC I am able to...

www.truenas.com

SOLVED - VMs can't see host

Hey, so I'm currently switching from TrueNAS 12.1 Nightlies to SCALE Nightlies. So far I only saw a few bugs here and there. But It does work. The only problem that I have right now is that my VMs can't see my host aka. TrueNAS. For example, my pihole-vm has the IP 192.168.188.21 and my...

www.truenas.com

For a system that advertises "virtualization" as one of main advantages (compated to TrueNAS core) it seems to be somewhat important functionality. Why do I still have this issue, two years after reporting? I was only able to get it work from boot menu when I connected monitor and keyboard to my truenas box and configured the network interface that way.

Somehing like that should work out of the box by default or be very easily done from GUI.
Instead it took me several days of research and experiments.

2. I am trying to make daily local replication form primary pool to backup pool on the same machine with TrueNAS Scale.
Again pretty basic and critical for a system that is primarily a NAS system, right?
And again there are many people that had this issue, some going again few years back:
https://www.truenas.com/community/threads/truenas-encryption-and-replication.93490/
https://jira.ixsystems.com/browse/NAS-109476
The only working solution seems to be to use the commandline as described here:
https://www.truenas.com/community/threads/truenas-encryption-and-replication.93490/post-647098
Having local replication from one pool to another seems to be a pretty critical and basic thing for a NAS.
Instead I am again spending more than a day on this issue, trying to get it work.

This all makes me wonder how the feedback between developers and users work in TrueNAS - how come we are having the same issues over and over again. That wastes time of so many people.

I really really want to love TrueNAS Scale.. Becase on paper it is exactly the system I want. ZFS with all its benefits and on top of that Linux virtualization and containers.
But.. I am starting to doubt my decision, it is very frustrating to accomplish these basic tasks for so long :(

 

[Arwen]

Was TrueNAS Core VMs able to see the host?

If not, then it's in essence, a new feature...

In general, production users don't use version 1.0 of any software. They wait, even up to a year, before starting validation of new software. I see great promise in TrueNAS SCALE, but many people think development is done and it's ready for all uses.

For example, when I read about TrueNAS SCALE's virtualization features, I think of Kubernetes not VMs.

 

[truecharts]

"This all makes me wonder how the feedback between developers and users work in TrueNAS"
File Jira tickets, this is (mostly) just the community forum not a "developers met-and-greet".

When working on software of any decent size, there comes a point that it becomes harder and harder to work on things without a ticket/issue system. It's also important to understand that each developer has it's own area of expertise, so even if a iX developer randomly stumbles on your post on the forum during his break, it would likely not be of interst to him/her. and surely wouldn't turn the guy/gall into an advocate for your specific issue.

And even if they do, some issues have a lot of consequences to fix or would be a huge time loss.
This means that other, possibly more important, things will get less attention.

But specifically to SCALE:
about half of the target features (it's aimed for hyper convergance after all, which isn't possible with it yet), are not even close to being finished.

 

[jeyare]

I think justifying the non-participation of developers is a futile endeavour.
They should not have much analysis of public voices. Otherwise, it is an inefficient use of resources. This is the first time I agree with anyone from TrueCharts here.

However, developers should have two additional layers on top of each other:
- SW analysts
and above them
- Business analysts.
Name it as you wish.

I think that these layers should be actively involved in:
- obtaining feedback from any sources. Jira is OK, but iX also needs a voice from future users. And where else but from the community?
- analysis of the feedback received
otherwise, it will just be a lab job for devs.
These are two fundamentally different roles that are negatively felt in the project results if they are missing or underestimated (capacity, knowledge, attitude). From a top point of view, this fundamentally underestimates the importance of the product.

I think that most of the shortcomings related to the Scale product were caused by the lack of clear information, e.g.:
- what we really do (web announcements vs deployment)
- for whom we do it or who is the customer (means paid customer) who will be a point for the return of investments. Because those who download it for free and don't buy iX HW or iX support can't balance it.
- when it will be really ready for use in production with the promised features.
- what is ready for a download now is just a "work in progress".
This is the Achilles tendon of the whole Scale project.

I keep reading here again (community) that what's on the web is actually completely different, and iX representatives here change the statements.
Last time here:

TrueNAS SCALE 22.02.1 Overview

There is a full blog here: https://www.truenas.com/blog/truenas-scale-22-02-1/ Release Notes: https://www.truenas.com/docs/releasenotes/scale/22.02.1/ Download: https://www.truenas.com/download-truenas-scale/ TrueNAS SCALE gets its First Major Update! TrueNAS SCALE 22.02.0 (“Angelfish”) was...

www.truenas.com

This is a reason why I follow this whole project only occasionally and I will try when it is really done. Because learning from Jira's answers that this isn't/won't be/it's not yet ... is a waste of time for me.

Don't get me wrong - but if it says on the official website that Scale is in RELEASE:
https://www.truenas.com/docs/scale/scalereleasenotes/
and then I read here (today) that:

about half of the target features (it's aimed for hyper convergance after all, which isn't possible with it yet), are not even close to being finished.

then:
iX should finally restart the principles of trustworthy communication.

I have sufficient experience creating similar systems, so I know that not everything is born easily. But there is too much chaotic noise around this project.

The last time I made a public statement about the difference between reality and the announcements, including instructions on how to resolve it, they locked the thread, and the second thread is already hidden for ordinary users - probably the notes still help them. Criticism is something that only a balanced personality can handle.

Five months passed, and I tried the latest version of Scale. Unfortunately, my opinion has not changed. See you in 5 months.
I cross my fingers for you. You have created a challenging goal.