rcrh

I'm planning a server upgrade and am trying to do it on the cheap. Demands on my server are pretty low. It delivers some media to a couple of Kodi installs and that's about it. My biggest driver is the number of drives I run. I currently have eight drives totally 19tb of storage. I'm currently running a 32bit pentium 4 processor with 4gig of RAM so just about anything I can buy today is going to move me ahead three or four years. The biggest bottleneck I find on my system is the PCi controllers for the extra drives. I currently have two controllers because I found running four drives off of one significantly impacted my performance. So I'd like to get all of the drives onto the mobo, be able to move a couple of server processes (emby or plex) onto the server, and reduce the time needed to run a parity check. asks: budget <$150 including RAM mobo with 8 SATA ports RAM at least 8gig I know I can pick up a Celeron g1840 for under $50 but I'm having problems finding an intel based mobo with the 8 sata ports. The closest I've come to finding what I want is this MSI 760GMA $55.00 Sempron 145 $45.00 My concern with this is that the Sempron 145 is a single core processor and just barely an upgrade while the next gen AMD processors start to stretch my budget. So in the end I think I'm looking for a suggestion for an LGA1150 socket mobo with 8 sata ports or a slap in the head for over emphasizing the ports. Thanks in advance.

garycase, ta very much.

Well aren't I an idiot!! Maybe some nice moderator will move this to 5.x/general support forum. Thanks again.

itimpi, thanks for the great reply. I appreciate all that info. Since the 3tb drive going in wasn't new I decided to start over so i could run a preclear on it. To do this I did the following: 1) stop the array 2) unassign the 3tb drive 3) run "new config" from the utility menu and wait for the parity to be rebuilt 4) run the preclear on the 3tb drive 5) add the drive to the array. I'm at step three at the moment and I'll take this opportunity to start reading about xfs. Although, I thought it wasn't available in version 5.x of unRAID. I thought it was a version 6.x feature. Thanks again.

First, thanks all. After reading this I got all kinds of cowardly and decided to get a little more manual in my upgrade. I was removing a 2tb drive so I cleaned up data until I had over 2tb available. Most of this space was gained by deleting things from the recycle bin. Next I manually moved data from the target drive to the other drives. Once it was empty I shut down the array, powered off the server and replaced the empty 2tb drive with the new 3tb drive. When I brought the server back up I select the new 3tb drive in that slot and expected unRAID to format the new drive & preclear it before coming back online. However, it didn't do that and just started a data rebuild. Eight hours later I now have what I wanted but am left with two questions: 1) the 3tb drive was NTFS formatted: did I miss the reformat or was it not done, and 2) should I be worried that the drive wasn't precleared? I've not put any data onto the drive so I'm at a perfect place to start over if I need to. Thanks again.

I'm running out of storage space and am about to upgrade one of the drives in my array. But I know that the initial preclear & rebuild of the data could take a couple of days. So I'm wondering if I can put the drive in a USB chassis, mount it, run the preclear, and then do the upgrade? I'm trying to minimize that period of time where I'm exposed to a failure without a complete back up. I don't think it matters but my config is: unRaid 5.0.6 processor: P4 of some sort RAM: 4gig drives: parity 4tb drive 1 - 4tb drive 2 - 4tb drive 3 - 2tb drive 4 - 2tb drive 5 - 2tb drive 6 - 2tb drive 7 - 2tb parity, 1, 2, & 3 are on the mobo controller and the others are on a PCI card. I'll be replacing drive 3 with a 3tb drive. Thanks in advance to anyone who can help me mount the USB drive & run the preclear script on it.

BRiT, that makes sense. I just assumed that the parity check would keep them all spinning for the entire process. With about two hours to go a fifth drive has spun down. However, I'm not sure why the smaller drives didn't spin down much sooner, like at about the half way point. I'm going to decrease the spin down delay and see what happens next time. Thanks.

I'm running a parity check (which for me is a 14 hour process :'( ) and I'm down to the last couple of hours and four of my drives have parked themselves. Is this possible? My array is parity 4gb drive1 4gb drive2 4gb drive3 2gb drive4 2gb drive5 2gb drive6 2gb drive7 2gb parity & first three are on the mobo controller and then the remaining are on two pci controllers (two on each). The four non-mobo drives are now parked. There is also significantly more free space on the four non-mobo drives. Does unRaid do something to determine when it gets to the end of data on drives? If not I would have expected all of the drives to keep spinning through the entire parity check. Thoughts? Thanks.

Is it possible to have all BIN files moved to a single drive? I have quite a mix of performance on my drives and I'm trying to maximize the space on my 'good' drives. I really like the functionality of the BIN but I'd be just as happy with having those files stored on one of my slower drives. I suspect I can just change the share settings but I didn't want to do something that could break the recycle bin. Thanks in advance.

Oh, so that's entirely unlike what I did. All of my drives with data went back into the array plus some new ones. Are you removing empty drives or are you expecting data to be rebuilt based on the existing parity?

the process above worked for me. I put in the new drives & the old drives; hit NEW CONFIG; then rebuilt the parity. The only thing that didn't work automatically was a rebalancing of the drives. I waited for the parity to finish so I had protection and then used the console & MC to move data from my old drives to the new 4tb drives. Last think I did was add my old parity drive to the array as a data drive. Not sure if this answers your specific questions.

they're in the rack of a coolmaster stacker case. The case has a chassis inside that holds four drives & has it's own fan. Like this. http://www.coolermaster-usa.com/product.php?product_id=2699

Thanks for that. I was using both. I'll go back and fix that. Is it better to "include" or to "exclude"?

I've just rebuilt my server and replaced a number of 2TB drives with 4TB WD Red and I'm surprised by the temperatures I'm seeing on the Reds. I thought on of the characteristics that made them a good choice for servers was their lower operating temps. During my initial parity build I was seeing temps in the upper 40c range (47~49) while my older 2TB seagate drives were about 10 degrees cooler. This leads to a couple of questions: 1) I'm I wrong about the temp expectation on the reds? 2) could something be reading the temp wrong from my reds? 3) is there any functionality that would log the temp of my drives over time so I can see if the rebuild process was simply a spike that I'll never have to worry about again? Thanks in advance.

Is it possible to control shares and drives so that a particular share is only written to a particular drive? I'm running eight drives across three controllers and there is a pretty significant performance drop off between my mobo drives and my PCi drives (PCix2 & just PCi). Because of this I would like unRAID to put my more archival material onto the slower drives. I have a share set up called ARCHIVE and I would like that share only written to my slower drives. Is this possible? Thanks in advance. **edit** I found my own answer right after posting this. It's right on the SHARES setting tab.

This is exactly my situation. I want the new drives and the old ones in the new configuration. I'm still slightly paranoid about data loss so I'm going to try this with one of my old drives first. If it works I'll add the other old drives and do it again. Thanks for this.

so I had some issues (and some spare cash) so I decided to upgrade my config. I've done the following: 1) bought three new 4TB drives, 2) downed my array and disconnected the existing drives 3) added the three new 4tb drives, 4) precleared the new drives, 5) added the new drives to the array (one as the parity drive). So I now have 8tb of storage. I would like to add back my old data drives. Do I a) have to mount them somewhere else & copy the data to the new array, or b) add them back to the array? If a), how do I mount them? within unraid & use MC to copy or under win7 & copy? If b), do I have to do anything special or will unRaid see the previously used unRaid drive and just build parity & balance out the storage usage all by itself? The only guide I've been able to find that talks about drives with data assumed the data is from win7 or some other non-unRaid OS. Thanks in advance.

Wow, this is great news. Where did you get the video breakout board? My MSS is still sitting around & I'd love to get it working.

Did anything happen with this? It would be helpful.

As compared to the glorious sound of the old school PC speaker! :D nudge nudge

Here's another voice in this crowd. I get this error whenever I have Plex running. Updates of the plex database seem to trigger the problem. I've looked at both the system log & the plex log and can't find any error messages. This is on rv11 of unraid and pms version 0.9.7.22.511 I have no other addons running. I'd be happy to test if you can give some guidance.

That's GREAT. I just found the "beep" command and had put a couple of random beeps into the go script. This way better. Thanks.

yeah, I noticed that too and I miss it. My server is in a different room and it would be nice to have the audible notification when the boot up is complete. Any way to get it back or add something to the go script that would do the same thing?

Thanks for this. I'll have a look at it in the AM.

Thanks Burt, I did remove SF and then reintroduced it. The server ran for three days with SF running and I didn't have a problem. The problems didn't return until I added SF AND Plex. Even then it took another four days before anything bad happened. I'm currently running without SF but with Plex. Plex is important while SF is something I like. We'll see how it goes. thanks again.