Can we keep the easy replication piece w/ FreeNAS Corral and get that into 9.x...THAT sure as heck was easier/a step in the right direction than in 9.x! :-D
FreeNAS Corral has a lot of great concepts that we'll be implementing in the next version. We've received other positive feedback on the new replication engine, so I don't see why we wouldn't pull that over.Can we keep the easy replication piece w/ FreeNAS Corral and get that into 9.x...THAT sure as heck was easier/a step in the right direction than in 9.x! :-D
That is great news...yeah I never could get peering/rep to work/happy between FreeNAS 9.10 to Corral...admittedly only spend 10-15 mins on it...there's always trusty zfs send/recv over a netcat tunnel thank goodness which is what I resorted to! :-DFreeNAS Corral has a lot of great concepts that we'll be implementing in the next version. We've received other positive feedback on the new replication engine, so I don't see why we wouldn't pull that over.
Thank you for being so open and honest. It is refreshing and I now understand the predicament you found yourselves in. I can surely understand it having been in a similar situation myself in my own working life as a software engineer.Fair questions, Pri. I'll have to walk a fine line between transparency and professional decorum here.
This is great feedback, STH, and a perception we're fully aware of. Over the years, and under previous leadership, the FreeNAS releases became what I liked to call a "QA crutch". It was easier to release, then fix, than it was to develop a strong internal QA framework for FreeNAS. All of that energy was instead put towards creating a stable enterprise product (TrueNAS), with all internal QA resources focused on that version of the software.Thanks for the comments Brett, I value the added clarity. Its a brave move stepping back from Corral and I can appreciate how difficult it must be for everyone at IX. My ask of IX would be to introduce some more stringent testing of the FreeNAS products going forwards, there were numerous issues with the '9' train which contributed to a general feeling of 'don't be the first to upgrade / cross your fingers' and I suspect a lack of internal QA also contributed to the Corral situation too. I understand FreeNAS is the free version of the commercial TrueNAS and part of the IX strategy is some QA is done by the FreeNAS users but I think its skewed too far now - certainly my confidence was declining before Corral and recent builds and ultimately this announcement doesn't do anything other than suggest I was right be wary.
Please take your time with the next 9 release and don't mess it up putting on the fancy lipstick.
If it is a dead 'code base' then I am bowing out as well and sticking w/ F5reeNAS 9.x for the time being. :-DNow that everyone knows they can safely roll back... anyone going to tinker with Corral and keep playing around with it?
....
Here's a link to the Jenkins test framework, if you're curious:
FreeNAS - Master - Install QA Tests [Jenkins]
Sorry, missed this post.I am in the middle of a very painful rollback at the moment. The upgrade failed due to one or another of the many known issues and only now do I find out that the project was just canned.
I can't believe it's still on the release server as STABLE. Why didn't they pull it down when they decided to cancel it? More bad product management or was leaving it up an intentional decision?
@brettdavis any plans to pull Corral STABLE off the release server?
I appreciate the understanding and support, Nnyan. Floating in a sea of bad options, we ultimately chose the one that was best for FreeNAS users, even though we knew we were going to displease a lot of them.When I first read the news about v10 I was very unhappy. Mostly b/c I was not a big fan of jails. But after reading here and other sources I have to admire the fact that pulling a product your not confident in while a very difficult and painful decision is the right one. That takes courage and yes it comes at a cost (in may ways, reputation, trust, community, etc...) it is by far the lesser of two evils. I know people are upset at time spent on v10 but such is life if you want to play with technology.
I switched about a year ago from a two system setup (pfsense + VMs on ESXi) and standalone Freenas to a converged Ubuntu server setup with KVM and ZoL. This has been working quite well and am now looking at docker/LXD to thin down the VM resource usage and minimise the modifications made to the host.I'll certainly be interested in alternative options for my home VM shared datastore.