TrueCommand 2.0 Nightly Images Now Available

kenmoore

TrueCommand Project Lead
iXsystems
Joined
May 1, 2019
Messages
51
The TrueCommand Team is pleased to announce that pre-release “nightly” images for TrueCommand 2.0 are now available for general testing!

TrueCommand 2.0 is a significant change from the 1.x series. The entire backend of the product has been re-created using Golang and InfluxDB rather than C++ and PostgreSQL. This means massive performance improvements, but also a major change to the product! We want to give version 2.0 an extended pre-release testing period before officially launching it.

For details about the new features coming in v2.0 or specific performance improvements, please take a look at our “Developers Notes” page.

TrueCommand Cloud Services also has an “opt-in” for customers to migrate their cloud instance over to the nightly releases. This updates your cloud instance to the latest “nightly” version every week. To opt-in to nightly versions with TrueCommand Cloud, contact iXsystems using the “Request Support” button in the iXsystems account portal. This is also available for any questions you might have about nightlies before opting in. Requests are confirmed by Digital Services representatives.

As this is a developers nightly image, all the usual precautions should be taken before testing, including backing up your previous data directory in case you wish to run the 1.3 branch again. This is not intended for production environments, only feature previews and testing.
Please DO create bug tickets if you encounter any issues while using a nightly!
 
Last edited:

nKk

Dabbler
Joined
Jan 8, 2018
Messages
42
I found a problem when try to connect TrueNAS-12.0-U2.1 system to TrueCommnad V2.0 (Master-20210225) docker.
If TrueNAS have HTTP to HTTPS redirection enabled TrueCommand return "websocket: bad handshake" and can't connect to TrueNAS.
When HTTP to HTTPS redirection is disable the connection working.
I create Jira issue TC-1602
 

ornias

Wizard
Joined
Mar 6, 2020
Messages
1,458
Nice work on those nightlies, I know it was somewhat of a thing to get nightly docker images being deployed nicely :)

I found a problem when try to connect TrueNAS-12.0-U2.1 system to TrueCommnad V2.0 (Master-20210225) docker.
If TrueNAS have HTTP to HTTPS redirection enabled TrueCommand return "websocket: bad handshake" and can't connect to TrueNAS.
When HTTP to HTTPS redirection is disable the connection working.
I create Jira issue TC-1602
Seems like either the user or Ken needs a little spanking for trying to use secure websockets on the http port.
 

sretalla

Powered by Neutrality
Moderator
Joined
Jan 1, 2016
Messages
9,702
First impressions: Wow! That's a great performance improvement over 1.x
 

sretalla

Powered by Neutrality
Moderator
Joined
Jan 1, 2016
Messages
9,702
And the new file explore feature... Totally unexpected, but genius!!! (of course we all know how to get to our files without it, but this makes it so much easier to get at things quickly from a single pane of glass.
 

sretalla

Powered by Neutrality
Moderator
Joined
Jan 1, 2016
Messages
9,702
At the moment it seems to be only exposing files and directories owned by root, so that's something to consider how to work with.

Also the breadcrumbs are a little odd in behaviour, but for an early look it's in great shape.
 
Last edited:

aervin

iXsystems
iXsystems
Joined
Jun 13, 2018
Messages
114
Yeah good point, there are some limitations but they can be overcome I'm sure. I'll ask the NAS MW devs about non-root access.

Breadcrumbs proved complicated for this feature, I'll admit. If you have time, please open a ticket with your ideas. I'll be happy to look them over with the team.
 

ornias

Wizard
Joined
Mar 6, 2020
Messages
1,458
@aervin Speaking of "non-root"...
It's a bit odd if iX is on one side building a Kubernetes based hyperconverged solution and on the otherhand TrueCommand can't run with non-root kubernetes securityContext settings... might be relatively important to look into! ;)
 

smillan

Cadet
Joined
Jan 15, 2021
Messages
8
hi,
where do i need to go to get the TC v2.0 beta as i want to upgrade my current TC 1.3.2 VM.

cheers,
 

Patrick M. Hausen

Hall of Famer
Joined
Nov 25, 2013
Messages
7,740
where do i need to go to get the TC v2.0 beta as i want to upgrade my current TC 1.3.2 VM.
 
Joined
Jan 4, 2014
Messages
1,644
Just installed 2.0 nightly. Very slick! My gripes with TC from 1.2.3 to 1.3.x have been addressed, namely:

1. System version information is back on the tiles
2. Better match of used/total space and corrected units (TiB).

TC just keeps getting better. I'm going to enjoy exploring the new version.
 

sretalla

Powered by Neutrality
Moderator
Joined
Jan 1, 2016
Messages
9,702
OK, I'm at the end of my patience...

I have tried for far too many hours now to set up a cluster volume on a second network I have in place behind my admin network...

I have 3 SCALE hosts on my 192.168.x/24 network on single Gbit and the same 3 hosts on a different, 192.168.y/24 5Gbit network as .27, .28 and .29 on both networks for each of the 3.

I have never been able to set up a cluster volume on the X network (1Gbit not enough, I guess).

Now I finally got around to putting the 5Gbit network in place, doing all the VLAN and bridge setup, I wanted to have a cluster volume... I'm not lucky today.

Every attempt I make to change something and then set up a volume, I get this:
Code:
NAS API error

"'IPv4Address' object is not iterable"


I have just worked out an issue that was blocking me (having thought it was a problem with having TrueCommand on one of the hosts) which also has some Traefik running on it... so I have redirected the standard ports to 81 and 444 respectively as suggested by the TrueCharts folks.

Could that be it? (that's all I can think of even though it doesn't even remotely resemble that message)
 

sretalla

Powered by Neutrality
Moderator
Joined
Jan 1, 2016
Messages
9,702
I believe the workaround is to use hostnames rather than IPs until the next RC release
Many thanks, will try that out.

It seems it's no good though:
System offline: dial tcp: lookup scale1.domain.internal on 172.17.0.10:53: no such host (well, that was in a Scale app, so using the kubernetes cluster DNS... now for another round with an external TrueCommand pointing at my router DNS)

OK, that's no good either... having set up all the hosts with hostnames from DNS on their admin ports, when I then set up the bricks, I am forced to select one of the 2 IP addresses for each brick... fine as far as I can tell as I don't want it to use the admin interface anyway... but we're back to square 1 with the IP address not iterable thing again.

Am I missing the point of what you said about using hostnames? I can't see where they would go to make it work differently.
 
Last edited:
Top