However if you have a separate volume for each disk, if the disk fails all data is lost. So if your data is important to you it is vital that you maintain a good backup.
The software on the Ultra is a port of an updated version of the software that shipped on the first ReadyNAS products released back in 2004. Before 2008 most ReadyNAS products had no more than 4 drive bays. Clearly adding support for more than 4 volumes wasn't seen as necessary. As the product is now EOL, it won't change in future versions of 4.2.x.
Having six separate volumes is not a good idea in my view as if any disk fails all data on that disk is lost. However the new OS, ReadyNAS OS does support creating a separate volume for each disk. So if you can backup all your data, you could update to that and restore your data from backup.
Yes, I am aware of the backup issues, thanks. This unit is being used as shadow storage on a private lan.
But, it should be trivial to support (number of logical volumes) <= (number of physical volumes). Seriously, we are talking about 6 vs. 4 volume objects for customer use.
I have worked for many years in operating system design and implementation, so I do have a firm grasp on the issues.
Also, the EOL issue is pretty bad, IMHO. The standard in the industry is 5 years of support after last ship before EOL. What Netgear is doing here is really unconscionable.
The standard in the industry is 5 years of support after last ship before EOL. What Netgear is doing here is really unconscionable.
Hmm... Netgear are still replacing out-of-production devices under warranty, they're continuing to update the OS4 and OS5 firmware and addons, their employees are offering informal support on this forum to owners of legacy devices, and they're even making changes to OS6 so it will work better on legacy devices even though it still isn't officially supported for those machines.
What terrible thing are they doing that I've missed?
The standard in the industry is 5 years of support after last ship before EOL. What Netgear is doing here is really unconscionable.
Hmm... Netgear are still replacing out-of-production devices under warranty, they're continuing to update the OS4 and OS5 firmware and addons, their employees are offering informal support on this forum to owners of legacy devices, and they're even making changes to OS6 so it will work better on legacy devices even though it still isn't officially supported for those machines.
What terrible thing are they doing that I've missed?
By not providing a SW upgrade path for customers who recently bought HW. I know that there are potential complications with upgrading, but those can be overcome. It is not an issue with the technology.
You say that like it's a bad thing. Clearly you haven't been reading the steady stream of OS6 complaints that have been posted here over the last year.
You say that like it's a bad thing. Clearly you haven't been reading the steady stream of OS6 complaints that have been posted here over the last year.
Actually, I am a newcomer to the forum. I actually came to ask a question, which has been answered. I will poke around more to see what is up with the new OS.
Actually, I am a newcomer to the forum. I actually came to ask a question, which has been answered. I will poke around more to see what is up with the new OS.
OS6 was a big change and was apparently rushed into production before it was ready, especially for the low-powered devices (the high-end x86-based boxes seemed to have fewer problems with the initial versions of OS6). Netgear has been working hard on the major showstopper bugs -- crashes and freezes, broken hardware, suboptimal filesystem configuration, incompatibility with the latest desktop operating systems, etc. -- and at this point those seem to have been fixed.
So that's good, but there are still plenty of less-critical problems. The web interface has basic flaws: obvious configuration options that are inexplicably missing, non-orthogonal backup options (to be fair, that's a problem in OS4 as well), seemingly trivial features like disk spindown that were available on OS4 but are still missing from OS6, etc. I've deliberately chosen not to run OS6 on my old boxes because OS4 still works better... At least for my definition of "better".
Anyway, welcome to the forum. I think you'll find a lot of interesting information here as you explore.
I've left my pro running OS4, but of course use OS6 on my RN102. I agree that the major showstopper type bugs have been fixed (including a couple in 6.1.7 beta).
OS6 is still missing some features I use - in addition to disk spindown, you cannot send a WoL packet in a backup job, and you cannot schedule disk scrubs and volume scans.
On the other hand, it includes some nice features that are missing in OS4 - including Antivirus, and more usable snapshots. It also will (sometimes) allow you to switch between flexraid and xraid w/o data loss.
Both the x86 OS4 and OS6 manuals clearly state that a 6-bay NAS can have up to 6 volumes. If that it not really the case, Netgear should either (a) fix it in a maintenance release or (b) revise the manual. (a) would be my preference, since I think users expect the ability to run jbod on any home NAS.
An alternative approach to getting more logical volumes would be iSCSI
I've left my pro running OS4, but of course use OS6 on my RN102. I agree that the major showstopper type bugs have been fixed (including a couple in 6.1.7 beta).
OS6 is still missing some features I use - in addition to disk spindown, you cannot send a WoL packet in a backup job, and you cannot schedule disk scrubs and volume scans.
On the other hand, it includes some nice features that are missing in OS4 - including Antivirus, and more usable snapshots. It also will (sometimes) allow you to switch between flexraid and xraid w/o data loss.
This is not encouraging about OS6. It sounds like it was released as beta quality, and they are playing catch up in the field.
Both the x86 OS4 and OS6 manuals clearly state that a 6-bay NAS can have up to 6 volumes. If that it not really the case, Netgear should either (a) fix it in a maintenance release or (b) revise the manual. (a) would be my preference, since I think users expect the ability to run jbod on any home NAS.
The limit is real. My guess is that someone in Marketing decided that they needed a 6 bay HW solution, but the software did not fully support it, and the company did not want to expend the engineering resources to fix the issues.
This limit is a real problem because you cannot even create a volume to test a new drive. You basically cannot do anything with the drive until you add it to an existing volume.
And, while we are at it, I hope that OS6 has better drive management support than OS4. OS4 lacks the ability to run targeted tests on drives, either offline, or using the SMART capabilities. Also, there is no way to wipe a drive which is being taken out of service. These are serious deficiencies.
The diagnostic information provided when a drive fails is basically non-existent. The NAS is a little computer with drives in it. It needs the ability to fully test, manage, and diagnose failures which occur. Requiring someone to have a PC available which they can plug a failing drive into is not a valid solution.
An alternative approach to getting more logical volumes would be iSCSI
I looked at iSCSI. It is an inadequate solution:
1) It is not a general file service. 2) Not all client OSes support it natively, often requires 3rd party software. 3) 2TB capacity limit is inadequate.
OS4 had the volume names c d e and f hard coded into the OS, so adding support for 6 volumes would require manually adding some lines of code. With OS6 you name your own volumes and it is written in such a way to allow more volumes than you are likely to need.
With OS4.2.x there is the disk test boot option, just like there is on OS6.
OS4 runs a short online SMART test daily at 4am. You can do offline SMART tests manually if you really want to (it's just not something that's likely to be explained here).
OS4 had the volume names c d e and f hard coded into the OS, so adding support for 6 volumes would require manually adding some lines of code. With OS6 you name your own volumes and it is written in such a way to allow more volumes than you are likely to need.
OK, but that is what software development is all about: "manually adding some lines of code".
With OS4.2.x there is the disk test boot option, just like there is on OS6.
That's pretty heavy handed. There is no way to verify a new disk individually or test a failing drive and getting any real diagnostics out of the system.
OS4 runs a short online SMART test daily at 4am. You can do offline SMART tests manually if you really want to (it's just not something that's likely to be explained here).
I have downloaded and examined the SMART logs on the unit. I see no evidence of this daily run. It looks like the short test is run once each time the disk is added to the unit.
As to running SMART tests yourself, I assume you are talking about using SSH? I thought that was not allowed by support?
OS4 certainly monitors SMART, I have gotten alerts when reallocated sectors and some other stats change.
I realize it monitors the SMART parameters, but I don't see that it runs any tests regularly.
On OS4 generally, it isn't getting new features, just maintenance updates. BTW, neither mdgm nor I work for Netgear.
I understand the status of OS4. However, they don't even seem to be qualifying new drives for the old configurations. That goes against the market in general. There are always new drives coming out, and old ones being discontinued. Soon, there will be NO drives on the market which have been qualified. What then?
I understand that you guys are not employees, and I appreciate your experience and insight.
