drawz

As noted above, the chipset is now integrated in that total 10W TDP. In addition, the latest Bay Trail CPUs support Speedstep, which the old versions did not. There is definitely some potential for reducing the real world power consumption *IF* the rest of the motherboard is designed with power consumption in mind. Even if it isn't, I still bet we see a decent drop.

nicinabox - i'm not sure what it is about boiler and the other tools you have developed, but they seem intimidating for some reason. your ideas are great and i was super excited about all of them when you announced them, including the Web GUI and boiler, but for some reason it feels like there is a barrier to entry that I think most users are not willing to go through. I think if you unified your GUI and package/plugin manager (which should run from the GUI primarily), you'd really make some progress. The lack of plugins availability is an issue too - it's a chicken and egg thing. Agree that limetech picking and recommending one solution would go a long way. For some reason, Tom seems to want to reinvent the wheel and only allow himself to develop anything official for unraid (see the webgui on github, which is a stripped down version of simplefeatures it seems).

Gee, I have both Dynamix and unMenu installed and I don't see the temperatures of any spun down disks on any of the screens that I looked at on unMenu screens. What screen are you looking at which shows temperatures of spun down drives? It only works on WD drives (I think) because they don't spin up when SMART data is requested. Your sig says you have Seagates and Hitachis. That's my best guess for why you don't see it in unmenu.

Glad to see this being released! Will give it a try as soon as we have UPS support! (this really should be built in!)

Whatever webGui is chosen, it all comes down whether emhttp stays alive or not ... Isn't it about time that all of the developers of webGUI's set the priority level of emhttp (I probably have the wrong terminology here as I am not a Linux expert) so that it does NOT get killed when resources get tight? I seem to recall seeing that it only takes a simple command that can be run from a script. I think that would be a function of emhttp and the OS and therefore up to Tom, not the webGUI developer.

That's what it's supposed to do. It is usually configurable. The only other way around this is to disconnect the USB or serial connection. The issue then becomes, when you come to the point of having to power down the machine because the batteries are close to depletion, there will be no communication. Not much risk to the server or data on the drives if it's in S3 and power is lost due to UPS depletion. You'll just have an unclean shutdown and possibly lose the state of any apps that are running. Your run time in S3 will be pretty long. Maybe it makes sense to disconnect the USB connection if you can't disable it in the BIOS.

Same here. I think this will be the best compromise of features. I'm just afraid that it will take forever to be labeled "stable" and/or there won't be a nice webGUI to administer it. Hoping that Open Media Vault takes care of the latter. Decided to see if there was any new info on BTRFS RAID 5/6 and came across the following post: http://www.phoronix.com/scan.php?page=news_item&px=MTU2NDQ and straight from the developers mouth: http://www.mail-archive.com/[email protected]/msg30103.html

Same here. I think this will be the best compromise of features. I'm just afraid that it will take forever to be labeled "stable" and/or there won't be a nice webGUI to administer it. Hoping that Open Media Vault takes care of the latter.

Same here. I think you need IPMI support to read the hardware monitoring data beyond CPU temp and I'm not sure if that works with the Dynamix system temp feature.

System Profiler/Info gives me the wrong CPU speed on my HP Microserver N40L. It think it is running at 1500MH instead of 800MHZ (and thinks the max speed is 2200MHz instead of 1500MHz). The info is correct if I telnet in and cat /proc/cpuinfo. This was a problem with simplefeatures as well, but now that there are fresh eyes on it, maybe we can find the issue. Obviously, this is essentially just cosmetic, although I'm not sure where else it would get the CPU speed from.

Thanks very helpful. I think the k10temp module is built in and doesn't need to be loaded via the go file. Haven't tried yet though. This should also work with Dynamix. I'd also be interested in getting this working! The only module available in the stock kernel will give you CPU temp. To do it; Put the following in /boot/custom/sensors.conf; # lm-sensors config chip "k10temp-pci-00c3" label temp1 "CPU Temp" Then add the following to the end of your /boot/config/go file; cp /boot/custom/sensors.conf /etc/sensors.d modprobe k10temp Reboot and your CPU temp should show. To get the other sensors we need to get the JC42 module found at http://khali.linux-fr.org/devel/lm-sensors/drivers/jc42/ compiled.

Had this problem on my Acer EasyStore H340. Never could find a solution. Shutdown fine on WHS and other linux distros. BIOS options were limited, but I tried all permutations related to power.

I think you have the plugin control in the wrong directory. Only the core webgui goes on /boot/plugins. The rest go in /boot/config/plugins.

It simply checks the presence of the progress file created by preclear in /tmp and displays its content. Ahhh good idea. Took a look in /tmp and see a bunch of other available info. Would be cool if you could pull some more of it in, like the SMART status after each phase.

Great idea - simple solution to this problem.

Thanks bonienl! Dynamix is a great addition to unRAID and, IMO, the best GUI for it yet. The plugin manager with integrated updating & changelogs is fantastic. I was also pleasantly surprised to see preclear status displayed on the preclear already running via screen.

Can't think of any reason not to update. Or at least try the update! Beyond my abilities to provide any useful testing though.

This article from arstechnica reminds us that security holes are everywhere and this is starting to be exploited on embedded devices, which we could consider unRAID to be in a sense. I haven't bothered to figure out which version of PHP unRAID uses, but I hope that 6.x brings us up to the latest Slackware releases and the updated packages that go with it. There are lots of good reasons to do this, including security and better plugin compatibility. http://arstechnica.com/security/2013/11/new-linux-worm-targets-routers-cameras-internet-of-things-devices/

This sounds awesome, but I'm not quite confident in my abilities to implement it (and don't have the time right now). Would make a great plugin with a nice webgui!

Seems doubtful, but could be a heavily fragmented disk, which might be more likely if full (and not filled sequentially).

It's always hard to know what's really going on when the sample size is small. And as usual with the internet, those with problems tend to be more vocal. I still don't see much feedback from those that tried it after MS had the opportunity to fix bugs and even fewer with R2. If your biggest complaint is the spin up/down issue, I think I can live with that because I will likely stick to only 3-4 drives. Like I said, I think I'll give it a try, possibly keeping unRAID around for media and WS2012E for backups and other server duties. Your virtualized solution is interesting, but I don't think it's one I'm willing to tackle right now.

People exploring unRAID for the first time, perhaps prior to trying/buying, could easily be mislead. Until Tom either officially announces an impending beta release or a beta is actually released, we probably should pretend it doesn't exist. It could be weeks, months, or years. Newcomers won't realize this though.

yes definitely. this one is especially confusing since there are several different scripts available

We probably shouldn't focus on btrfs coming to unRAID too soon. Given that it hasn't even been seen in beta or RC form, it's probably a long way off given Lime Tech's history. Let's not give anyone false hope or let someone buy unRAID thinking a certain feature is just around the corner.

Having all the drives in a storage space spin up is expected behavior - it's necessary by the nature of being striped since the data for a single file is spread across disks. This should be a performance advantage since the data is coming off multiple drives in parallel. On the other hand, unRAID's setup without striping allows individual drives to spin up or be removed from the array and still be readable. I did visit that forum and was unimpressed by the relatively small number of bad experiences. In fact, there weren't that many experiences overall and many of them seemed to be based on hearsay, jumping on the band wagon, or simply based on old information (ie immediately after WS2012 release a year ago). I'm still tempted to give it a try with a few spare drives. It does sound like there are some limitations, but every solution has limitations or we'd all be using the same setup.