TrueCommand 2.2 Makes its Debut

[wsoteros]

TrueCommand 2.2 is now available as a “Nightly” image for testing, ahead of its planned release over the coming weeks in mid-July 2022. TrueCommand 2.2 is necessary to make use of the additional TrueNAS SCALE 22.02.2 APIs, specifically for Clustered SMB.

TrueCommand 2.2 is another major step in improvements for TrueCommand. There are over 170 bug fixes, improvements, and features. The most notable changes are:

Reworked Stats engine: Moving from influx to rrd files makes TrueCommand more robust if memory is in short supply. This will also make TrueCommand Cloud more reliable for larger deployments. TrueCommand 2.2 will automigrate the configuration and stats when it first boots up. A longer QA cycle is appropriate given this major change.

Improved Certificate handling: Identified and resolved a number of issues with Certificates that should simplify deployments.

Clustered SMB support: APIs and a wizard for setting up an SMB cluster are included. These can be used with TrueNAS SCALE 22.02.2 (or later).

Improved Middleware: Better error handling and testing frameworks will ensure quality in the future.

TrueCommand has been well tested with TrueNAS 13.0 as well. TrueNAS 13.0-U1 has scalability improvements which ensure larger systems can be safely deployed with TrueCommand.

The nightly version of TrueCommand 2.2 is available on DockerHub with the tag “nightly”. TrueCommand remains free for systems with less than 50 drives.

The nightly version should only be used on test systems and not for production. The supported version will be announced and available in mid-July after the quality assurance benefits of internal and Community testing. Thanks in advance to all those that will test and provide feedback.

 

[justsomeguy]

Glad to hear it. I noticed the nightly was updated within the past hour, so we'll probably test it this afternoon. Right now we're running 2.1.1 which connects to an M60-HA. We encountered some bugs/issues with TC 2.1.1 and as such we're not running it in production yet. Some of those bugs are supposedly fixed in 2.2, so we'll see how it goes.

 

[justsomeguy]

The migration/upgrade worked well. Settings seemed to transfer over with no problem. We were finally able to configure unauthenticated SMTP for alerting, so that's good!

Still struggling with creating an AD/LDAP(S) connection. Error message from server: "...LDAP bind without requesting signing (integrity verification), or performed a simple bind over a clear text (non-SSL/TLS-encrypted) LDAP connection."

Not a new problem though - same as we were getting on 2.1. Probably something misconfigured on our end. We have support through iXsystems, so I'll put in a ticket tomorrow to see if we can get some help with that.

 

[justsomeguy]

I logged in this morning and the system we were monitoring looked like this. Had been showing up just fine for weeks in 2.1.1. Tried deleting (to attempt adding it back), but after waiting a few minutes it never deleted. Went ahead and rolled back to a snapshot from yesterday and it's working as expected again. I figure we needed to roll back to a production version anyhow before submitting a ticket for the LDAP issue above. Can attempt to recreate later on if needed.

 

[morganL]

I logged in this morning and the system we were monitoring looked like this. Had been showing up just fine for weeks in 2.1.1. Tried deleting (to attempt adding it back), but after waiting a few minutes it never deleted. Went ahead and rolled back to a snapshot from yesterday and it's working as expected again. I figure we needed to roll back to a production version anyhow before submitting a ticket for the LDAP issue above. Can attempt to recreate later on if needed.

Not what we want... please "report a Bug" or open a case with our support team.
If its a common issue we'll probably find with our testing, but if its less common we might need help finding it ourselves.

 

[Astraea]

I am not sure this is the right place to be making this post but this afternoon I spun up a set of 3 TrueNAS Scale servers in a VM to test out the new version of Scale and TrueCommand. So I also spun up a new version of TrueCommand on my laptop using the nightly build. Once everything was installed and configured as I normally would for a new Scale install, I was able to create a cluster but when trying to complete the cluster setup for the SMB setup. On 1 of the servers I could not select the IP address and on the other 2 I could but on all of them I could not select the interface and under the Interface and Static IP options, there was an error 2 for all of them. I did shut all three Scale installs off and added a second virtual NIC that only the three nodes used for cluster communication and also restarted TrueCommand. After that, I was able to get past the network setting screen.

However I ran into another issue, it wants to only continue using Active Directory configured as that is apparently the only way it supports as there is no way to sync SMB users across the cluster. I only use Linux in my setup so I do not have Active Directory configured so I cannot move past that screen. Guess I am back waiting for another update to TrueNAS and or TrueCommand before I try again.

Does anyone know if there will be alternative methods for cross-cluster user management such as LDAP or even just an option in the GUI to create/manage cluster-based used that get created on each node in the cluster with the same UID and GID. I did find in some documentation that users with the same UID and GID will work as long as they are on each node but I was not sure how to complete the setup in TrueCommand using this method.

 

[morganL]

I am not sure this is the right place to be making this post but this afternoon I spun up a set of 3 TrueNAS Scale servers in a VM to test out the new version of Scale and TrueCommand. So I also spun up a new version of TrueCommand on my laptop using the nightly build. Once everything was installed and configured as I normally would for a new Scale install, I was able to create a cluster but when trying to complete the cluster setup for the SMB setup. On 1 of the servers I could not select the IP address and on the other 2 I could but on all of them I could not select the interface and under the Interface and Static IP options, there was an error 2 for all of them. I did shut all three Scale installs off and added a second virtual NIC that only the three nodes used for cluster communication and also restarted TrueCommand. After that, I was able to get past the network setting screen.

However I ran into another issue, it wants to only continue using Active Directory configured as that is apparently the only way it supports as there is no way to sync SMB users across the cluster. I only use Linux in my setup so I do not have Active Directory configured so I cannot move past that screen. Guess I am back waiting for another update to TrueNAS and or TrueCommand before I try again.

Does anyone know if there will be alternative methods for cross-cluster user management such as LDAP or even just an option in the GUI to create/manage cluster-based used that get created on each node in the cluster with the same UID and GID. I did find in some documentation that users with the same UID and GID will work as long as they are on each node but I was not sure how to complete the setup in TrueCommand using this method.

Probably worth a separate discussion thread.
1) yes there needs to be a separate back-end and front-end network for a cluster
2) AD is the default way of coordinating SMB nodes.