Heads up! 21.06 BETA out soon

[Kris Moore]

SCALE testers,

Just a quick heads up. We're scheduled to release the next major SCALE milestone, 21.06-BETA on June 15th. At this point we've reached code-freeze and are busy working on fixing all the bugs and regressions we can find before this is tagged and goes out the door. If you are interested in kicking the tires early, please do so by grabbing a nightly image, and then report any and all issues you find to https://jira.ixsystems.com

Thanks and looking forward to a successful BETA launch cycle!

 

[ornias]

Thanks for the headsup @Kris Moore

TrueCharts has also done some major rework/fixes for 21.06 BETA. :)

Any issues for which can (considering the short timeframe till the Beta release) be directly send to the support channel on our Discord, that way we can get them fixed in time for Beta release.

 

[leckyBill]

Is it going to be an update through the UI or will this require a backup and new install?

 

[ornias]

Is it going to be an update through the UI or will this require a backup and new install?

Through the UI :)

 

[Trexx]

If you are interested in kicking the tires early, please do so by grabbing a nightly image, and then report any and all issues you find to https://jira.ixsystems.com

Thanks and looking forward to a successful BETA launch cycle!

As a relatively new user of TrueNAS (but have had personal NAS at home for many years, as well as working with large scale enterprise NAS platforms), the SLOW performance of the Jira platform really dis-incentives the submission/updating of bugs. When it takes almost a full minute or more just to log Into the jira platform, just to help document issues, that really doesn’t create a positive user impression.

I am hoping that the Jira platform receives a HW upgrade/DB optimization/etc. (whatever is causing the poor performance) to help make the submission process less painful.

I look forward to the release of the 21.06 beta in less than 2 weeks.

 

[stavros-k]

As a relatively new user of TrueNAS (but have had personal NAS at home for many years, as well as working with large scale enterprise NAS platforms), the SLOW performance of the Jira platform really dis-incentives the submission/updating of bugs. When it takes almost a full minute or more just to log Into the jira platform, just to help document issues, that really doesn’t create a positive user impression.

I am hoping that the Jira platform receives a HW upgrade/DB optimization/etc. (whatever is causing the poor performance) to help make the submission process less painful.

I look forward to the release of the 21.06 beta in less than 2 weeks.

A minute? That's fast! I've seen 20 min :D Anyway, it's already discussed, and the slowness is because of the location. iX is getting ready / planning to migrate Jira to cloud with servers all around, so that will be fixed!

 

[Kris Moore]

A minute? That's fast! I've seen 20 min :D Anyway, it's already discussed, and the slowness is because of the location. iX is getting ready / planning to migrate Jira to cloud with servers all around, so that will be fixed!

Yep, we're working towards a migration since we know how slow the server version can be. Just takes time like everything else :)

 

[stavros-k]

Yep, we're working towards a migration since we know how slow the server version can be. Just takes time like everything else :)

Yea, I still prefer you spend most of the time to truenas development (because, in the end, this is the end product!) than jira!

(Btw, I found that using a VPN and connect to US server helps most of the time. I manage to browse Jira, when I could not even login without the VPN).

 

[Gcon]

Just downloaded and installed the most recent nightly build. Wow wow wow - I love what you've done to the menu system. So clean! Just waiting for some parts to arrive for a test build on a refurbished Dell PowerEdge R710. Wil be testing iSCSI, NFS, SMB, SNMP, FTP, WebDAV, TFTP, SSH, LLDP, UPS. KVM (Nakivo B&R on Ubuntu), Plex docker, bunch of other dockers, dual M.2 SATA ZFS on root, SMART, scrubs, rsync and more. Will report in any issues. Thanks for the great work thus far.

 

[j_r0dd]

ACL's?!?! I spent entirely too much time getting those fine-tuned on CORE and dread having to do all over again in SCALE

 

[Yorick]

ACL's?!?! I spent entirely too much time getting those fine-tuned on CORE and dread having to do all over again in SCALE

I am feeling unsure. https://jira.ixsystems.com/browse/NAS-104325 seems to suggest yes but also "Kernel / ZFS changes still pending. " So ... not quite yet? [NAS-103664] Add support for NFSv4.X ACLs on Linux - iX - Bug Tracker (Jira) (ixsystems.com) remains open.

 

[beagle]

I am feeling unsure. https://jira.ixsystems.com/browse/NAS-104325 seems to suggest yes but also "Kernel / ZFS changes still pending. " So ... not quite yet? [NAS-103664] Add support for NFSv4.X ACLs on Linux - iX - Bug Tracker (Jira) (ixsystems.com) remains open.

There is a post somewhere in the forum saying it depends on some changes being implemented upstream.

 

[j_r0dd]

I am feeling unsure. https://jira.ixsystems.com/browse/NAS-104325 seems to suggest yes but also "Kernel / ZFS changes still pending. " So ... not quite yet? [NAS-103664] Add support for NFSv4.X ACLs on Linux - iX - Bug Tracker (Jira) (ixsystems.com) remains open.

Yeah I found the same on Jira. Looks like it’s mostly in place. For my use case it’s my only hold up for not migrating yet. Other than that everything has been solid in my VM except some bugs introduced in some nightlies which is expected when living on the edge.

 

[anodos]

Yeah I found the same on Jira. Looks like it’s mostly in place. For my use case it’s my only hold up for not migrating yet. Other than that everything has been solid in my VM except some bugs introduced in some nightlies which is expected when living on the edge.

Backend stuff has been merged, but migration pathways have not been tested yet. This is an area where beta testing will be helpful, but once again it's not a great idea to alpha (or beta) test with your production data.

There is still work needed to make Ganesha aware of native NFSv4 ACLs, but middleware APIs and Samba should work.

 

[j_r0dd]

Backend stuff has been merged, but migration pathways have not been tested yet.

Is it in the latest nightlies? If so ACL Manager is still not working in the gui, but I don't exactly need it to be working in the gui yet. Just need my perms to still exist.

This is an area where beta testing will be helpful, but once again it's not a great idea to alpha (or beta) test with your production data.

I've been using the alpha in a VM for a couple months now to familiarize myself with k3s and charts without anything devastating happening thus far. Maybe i'll try loading this weekend and i'll report back on ACL's.

 

[anodos]

Is it in the latest nightlies? If so ACL Manager is still not working in the gui, but I don't exactly need it to be working in the gui yet. Just need my perms to still exist.



I've been using the alpha in a VM for a couple months now to familiarize myself with k3s and charts without anything devastating happening thus far. Maybe i'll try loading this weekend and i'll report back on ACL's.

You'll need to make sure acltype on dataset is nfsv4. CLI management tools are nfs4xdr_setfacl, nfs4xdr_getfacl, and nfs4xdr_winacl. output / input should be same as FreeBSD commands. setfacl is easiest in the interactive editor (vi) nfs4xdr_setfacl -e <path>. Once you've set, recursive changes can be applied via nfs4xdr_winacl -a clone -rv -p <path>. JSON input / output is also supported for the CLI tools.

 

[j_r0dd]

You'll need to make sure acltype on dataset is nfsv4. CLI management tools are nfs4xdr_setfacl, nfs4xdr_getfacl, and nfs4xdr_winacl. output / input should be same as FreeBSD commands. setfacl is easiest in the interactive editor (vi) nfs4xdr_setfacl -e <path>. Once you've set, recursive changes can be applied via nfs4xdr_winacl -a clone -rv -p <path>. JSON input / output is also supported for the CLI tools.

Is there any reason the usual binaries getfacl or setfacl should not be used?

 

[anodos]

Is there any reason the usual binaries getfacl or setfacl should not be used?

They are designed for POSIX1E ACLs.

 

[oumpa31]

If I'm reading this correctly the ACL permissions aren't working yet in the GUI. So moving from Core to Scale permissions are still up in the air if you have ACL's set in Core its not recommended to switch over to scale.
I know I have to create all the Users and Groups over again when I move to Scale because I plan on doing a fresh install. Would the best thing to do when i get things ready to move is to remove all the ACL permissions at this point or should i wait until 21.08 comes out. I was really hoping to get Scale on my main server at 21.06 so my plex transcoding can be done with a video card instead of the CPU.

 

[cyrus104]

What about a fix for those of us that have already moved from Core to Scale and are having ACL permission issues? Is there a script that be applied? I really just need to wipe out all ACLs and apply new ones.