Hi All,

 

After about 3 years on my first build i've had no real issues. It's been a rock solid media file storage server solution. Now however i'm moving into a larger house with multiple instanced of xbmc and would like my server to power all (3) simultaneously.

 

I'm not interested in Passthrough graphics as i prefer to run thin clients (HTPC's) to power graphics/sound on each individual XBMC instance. I dont use newsgroups for downloading either. I do use utorrent and atm have a dedicated netbook that sits atop my unraid box that handles downloading/renaming/caching. I would prefer to eliminate the need for the netbook and would like my next unraid server house a centralized XBMC library on a SSD. I believe this requires MySQL and MySQL is best setup on virtualization? Virtualization would also offer me the option of eliminating the netbook by allowing a place for utorrent.

 

So my question is which one of Tom's prebuilt all-in-one server solutions (AVS 10/4) would you guys recommend? Obviously the Atom build is too under powered for my needs. But would i need the Quad core? 16 or 32 GB of RAM?

 

Second question, I'm interested in having Limetech configure a few drives for me. So i would purchase 1 HDD and 1 SSD drive and have them preconfigured as the parity and cache drives. However the order form doubles the shipping price when i add HDD's to my order. Im in Australia so the shipping price gets ridiculous really fast. I've read elsewhere here that the shipping price shouldn't change much as long as the weight stays below 75 lbs. Can Limetech setup a custom order form for me?

 

Third question, im still using unraid Plus (4.7) as i've never upgraded the usb key. I will need to upgrade the to 5.xx ? Additionally i will purchase a Plus to Pro upgrade for this new server, but wont Limetech need the physical USB drive to preconfigure my server? What is the solution here? Will I be forced to purchase a new Pro license?

 

Any advice will help.

 

Cheers,

Steve

 

 

I'd go with the Xeon to ensure you have ample processing power for your VMs.  16GB should be fine ... you can always upgrade the RAM later if necessary, but I doubt you'll find that to be an issue.

 

Tom will need to weigh in on whether or not adding a couple drives will really necessitate the higher shipping costs -- if so, I'd just buy the drives locally.

 

Note that a USB flash drive and Pro license is included in the price of the server.  I assume this is setup with the latest v5 stable release -- NOT with one of the v6 Beta releases that have built-in virtualization support.    You may want to just run v5 until v6 stabilizes;  then upgrade to v6 and setup your VMs to eliminate a couple of your other systems.

 

 

Great, thanks for that Gary, any idea if v6 is closer to 6 or 12 months from stable release?

 

I'll wait for Tom to have a look at the order (attached) I would be willing to spend approx $400 on shipping.

Great, thanks for that Gary, any idea if v6 is closer to 6 or 12 months from stable release?

I strongly suspect that not even LimeTech know the answer to this   However now that 2 new members have joined LimeTech it is likely that progress should be faster than in the past and I am sure that must be LimeTech's aim for justifying paying for these new resources.  Having said that if you simply use the existing v6 beta's in basic mode without trying to exploit the new virtualisation features most people are already finding it to be stable enough for day-to-day use on live servers.

 

We have already been told that for a forthcoming beta LimeTech will almost certainly be moving the virtualisation support from being Xen based to being KVM and Docker based.  Until this is released to the rest of us we have no idea whether it will affect v6 stability or what it will do to the v6 timescales.  It is always possible that new issues will appear with these virtualisation products that may lead to another rethink on the best way forward.

Agree that it's virtually impossible to predict the time schedule for v6 final -- it's an ambitious release; but should be a VERY nice product ... rock solid NAS functionality with built-in support for virtualization.

 

If you haven't already, you should PM Tom r.e. your order and the shipping costs.

 

Great, thanks for that Gary, any idea if v6 is closer to 6 or 12 months from stable release?

 

You need to define "stable release".

 

The focus of 6.0 is on virtualization and container technology. The current beta (5a) is quite stable and usable having gone through 5 beta releases. It is important that you pin cores to the unRaid "dom0" and you are good to go. Ironicbadger's excellent VM is a fantastic piece of work. I am running 5a and love it.

 

But Limetech is looking at some different VM tools including KVM and Docker due to hitting some issues with passthrough of video with Xen. So I would not recommend those interested in video passthrough to use it on 5a unless they are made of study stock. There may be one more Xen beta but I think not.

 

The next beta cycle will begin with b6 or b7. It wil be the first beta with the new toolset. I think it likely we'll see a series of quick betas in which the kinks are worked out and we'll get to a stable beta, similar to the maturity we see now in 5a and many users will migrate to it. In the past we've seen a series of very specific issues with mature betas that affect specific hardware configurations and ultimately need a new Linux release. They have nothing to do with unRaid itself. And we can see a beta sit while Tom waits for the issue to be fixed. I call these stable betas and have and will recommend users to migrate to them as their maturity increases.

 

Realize that the difference between beta and release is not that great. Releases often have small defects that are addressed over time. We are not like Windows with millions of users and formal beta cycles involving thousands and thousands, maybe more, users kicking the tires. We do not know the exact number, but we may be lucky to have a couple dozen folks with test arrays to try out early betas. But if more users don't migrate until the betas become final releases then unRaid will wait until the final release to be well tested, and that's where the bugs will be found.

 

So I encourage us to not look at the final release as some sort of magical milestone. Follow the beta progress and when you see one that, after a month or more of testing in the community has occurred, and there are few reported issues (that are not applicable for you), that you consider taking the plunge into that release. You will benefit from new features much faster.

 

One final comment - the core of unRaid, the part that maintains parity, is not undergoing modification. In this regard the earlier 5.0 betas were more "dangerous" as we were moving to GPT partitions that maybe touched the core of unRaid (and even then no beta I can recall had a bug in maintaining parity). If that part of unRaid is rock solid then your data is safe. If you feel safe about your data a few bugs in specific features that you don't plan to use right away, or related to a hardware configuration quite different from your own, may not be a reason to stay behind.

 

Another point, sticking to a Limetech server or one that closely mimics one is a great way to avoid some of the obscure bugs we sometimes see with specific hardware. You know Limetech will be testing with the servers they sell.

 

Thanks for that guys, and i def agree with that last point. Toms all in one servers are reasonably priced. I'm not seeing a lot of markup there. I'll have to do some more reading on here and get myself back up to speed.

Great, thanks for that Gary, any idea if v6 is closer to 6 or 12 months from stable release?

 

I'll wait for Tom to have a look at the order (attached) I would be willing to spend approx $400 on shipping.

As a person who builds and sells unRAID servers that shipping price is not to bad really.  A fully insured server shipped across the pond can run you near $800.  I am currently building 2 that are going to Thailand and shipping is not cheap for sure.

 

I have used Shipito as a intermediate stop.  Basically it allows us, the builder to ship and insure for a local US shipment, and then from there the buyer works with Shipito to get it to it's final destination.  This usually ends up working out a bit cheaper then a direct shipment across the pond.  The only way you get good prices for shipping over seas is if you are shipping pretty large volumes and I know I don't qualify for that sort of treatment.

I have checked out ShipitTo and the others but would prefer it shipped direct from the supplier. I have lost items before using re-shippers, thanks for that though.

 

Plus local shipping + assembly + au markup on parts cancels out most of the shipping cost.

I have checked out ShipitTo and the others but would prefer it shipped direct from the supplier. I have lost items before using re-shippers, thanks for that though.

 

Plus local shipping + assembly + au markup on parts cancels out most of the shipping cost.

You also have to take into account customs fees.  Most of my customers "forget" about that part until I mention it.  I do not cover customs fees, as I leave that up to the customer but depending on where it is going it can be a little chunk of change.

 

I have had a few customers ask about declaring the value of the shipment to be considerable less than what it is, that is never a good idea for the seller.  We have no way of recouping the cost of the build when we do that and if something should happen during shipment.  I have only every had one server arrive over seas dented and dinged, thankfully since it was covered fully the shipping company took care of covering all the replacement parts and shipping.

Good point it will be over $1000 so if customs catch it they will ask for some money. I can't see them taxing me in the shipping costs or the costs up to $1000 so it shouldn't be much.