One2go Posted January 18, 2010 Share Posted January 18, 2010 Like many others here, I was searching for a motherboard that was not old but actually could be bought from on-line retailers. After searching the forum here and checking with a few other sites I decided to give the ASUS M4A78-E motherboard a try. I trusted that the AMD chipsets would work from the different comments that I researched and sure enough I believe this mobo could be a winner. The cheapest on-line retailer was Zip Zoom Fly but they were out of stock so Newegg it was. Here are the details: North Bridge AMD 790GX South Bridge AMD SB750 2 x PCI Express 2.0 x16 (Single x16 or Dual x8) 2 x PCI Express x1 SATA Ports 3Gb/s 5 Onboard Video Chipset ATI Radeon HD 3300 GPU Video Ports D-Sub + DVI HDMI 1 x HDMI Max LAN Speed 10/100/1000Mbps USB 1.1/2.0 6 x USB 2.0 IEEE 1394 1 x IEEE 1394a eSATA 1 x eSATA 3Gb/s Form Factor ATX I installed the board with an Athelon X2 (AM2+ Socket), 2 Gig of OCZ DDR2-1066 Mhz that I had, Corsair TX750 PSU, Antec 1200 case, 4Gig Patriot Memory USB key and 4 x Samsung 2TB hard drives to round off the unRAID server components. I followed the instructions for installing the unRAID OS and powered up the Tower. After making changes in the BIOS, like booting of USB key, memory voltage 2.1 to get the OCZ to run at 1066 MHz plus AHCI mode for the Hard drives I rebooted the server and voila we are go. After more reading and consulting I installed the preclear script and I am almost done with burning in the drives. After 30:54:55 we are 76% complete of the post-read process. Next step for me is creating the shares and I guess doing some more tests to see if this will become an excellent motherboard. With the PCI-e slots available and the on-board SATA ports the potential is to have 17 drives hooked up. Any suggestions on where to go from here to make sure that this motherboard will measure up to the requirements set by the other reference boards? Link to comment
batfink Posted January 18, 2010 Share Posted January 18, 2010 Any suggestions on where to go from here to make sure that this motherboard will measure up to the requirements set by the other reference boards? To get started see here for a thread on motherboard testing http://lime-technology.com/forum/index.php?topic=3272.0 and here for MB compatibility info http://lime-technology.com/wiki/index.php?title=Hardware_Compatibility#Motherboard Cheers Link to comment
One2go Posted January 18, 2010 Author Share Posted January 18, 2010 I did read the compatibility chart and that is where I got some ideas of what mobo to go for from. At present the server is preclearing four 2TB Samsung drives and it is taking quite some time. To make it worst after 30 hours and being 76% into the post read process my PC with the 4 telnet sessions open decided to restart after a windows update, will not happen again. So at present I am going for the second time through a preclear process. Damn 2TB is a lot of data to read and write to. Once I have the 4 drives on-line I'll try to follow up on the level 1 test and once the shares have been assigned and content has been copied I should have some idea how compliant this motherboard is. On a side note, how do you start off with a clean syslog and when running a preclear from the console will activity be logged in the /var/log/syslog or where do I look for logged reports? It has been a very long time for me since fooling around with VAXs and using vi. Link to comment
batfink Posted January 18, 2010 Share Posted January 18, 2010 I'm a bit of a noob myself so I'll defer to the (many) more knowledgable bods on here, but if you haven't already done so install unMENU which will make capturing a syslog much easier. Re: 2TB HDD, tell me about it.... Just got a new 2TB and my PC locked up on me...... now re-running. 25 Hours, 48 mins and counting........ Link to comment
One2go Posted January 19, 2010 Author Share Posted January 19, 2010 My Samsung 2TB will take about 31h to 32h to finish all the tests. But I'll say better safe then sorry. It's a whole new experience to get back into unix syntax. As once a wise man said Unix was invented by space aliens. Link to comment
cj0r Posted January 26, 2010 Share Posted January 26, 2010 You should install the screen pkg that way you don't have to keep your telnet/putty sessions open on your desktop and you won't ever have that unfortunate preclear accident ever again. Link to comment
One2go Posted January 26, 2010 Author Share Posted January 26, 2010 I looked at that but the linux newb I am decided putty was easier Here are my entries from the log: Minor issues: Jan 26 18:22:52 Tower kernel: ACPI Warning: Incorrect checksum in table [OEMB] - 7D, should be 74 20090521 tbutils-246 Jan 26 18:22:52 Tower kernel: ata1: softreset failed (device not ready) Jan 26 18:22:52 Tower kernel: ata1: applying SB600 PMP SRST workaround and retrying Jan 26 18:22:52 Tower kernel: ata4: softreset failed (device not ready) Jan 26 18:22:52 Tower kernel: ata4: applying SB600 PMP SRST workaround and retrying Jan 26 18:22:52 Tower kernel: ata2: softreset failed (device not ready) Jan 26 18:22:52 Tower kernel: ata2: applying SB600 PMP SRST workaround and retrying Jan 26 18:22:52 Tower kernel: ata3: softreset failed (device not ready) Jan 26 18:22:52 Tower kernel: ata3: applying SB600 PMP SRST workaround and retrying Errors Jan 26 18:22:52 Tower kernel: atiixp 0000:00:14.1: simplex device: DMA disabled Jan 26 18:22:52 Tower kernel: ide1: DMA disabled Jan 26 18:22:53 Tower kernel: xor: automatically using best checksumming function: pIII_sse Any comments on the syslog. I guess since I don't use IDEs the DMA disable doesn't matter. syslog-2010-01-26.txt Link to comment
cj0r Posted January 27, 2010 Share Posted January 27, 2010 I'm a linux nub too you use screen through putty though easy as pie once you install it. http://www.oreillynet.com/linux/cmd/cmd.csp?path=s/screen there's a guide to the commands etc. Basically you just need to know how to connect/disconnect to a screen session and how to close them when you're done. Link to comment
One2go Posted February 4, 2010 Author Share Posted February 4, 2010 Copied 185 Gig to the array and did my first parity check. Accordingly to the log it took 7h & 45 minutes but accordingly to my watch only 7h & 30 minutes. Well 15 minutes is not the world. Will the parity check time increase as the array gets fuller? So far I am very happy with the combination of hardware no hiccups unless someone else can spot them in the Syslog. I searched and researched the forum and wiki for a simple guide for setting up for the first time the shares, or an example of when finished building the box and the array is on-line what next in Linux newb terms. I ended up creating the User Shares Bits and Bobs, Documentaries, Movies, Pictures and TV Shows. I wanted to have Movies & TV Shows on Disk2 & 3 while the others on Disk1. But the disk shares showed that all the directories where created on Disk1. Eventually I deleted and manually created the directories on the Disks, then the User shares worked with the parameters Fill-Up, Free Space=31457280 (30Gig), Include Disk1, Exclude Disk2,Disk3 and vice versa for the Movies & TV Shows shares. Any idea why that is the case? Spent too long of a time trying to figure out how to use smbmount to copy from a QNAP NAS to the tower. Did it via windows explorer & TeraCopy. Got between 28 & 40 M/sec. Maybe Next time I try smbmount. Feb 4 01:45:29 Tower kernel: mdcmd (856): check CORRECT Feb 4 01:45:29 Tower kernel: md: recovery thread woken up ... Feb 4 01:45:29 Tower kernel: md: recovery thread checking parity... Feb 4 01:45:29 Tower kernel: md: using 1152k window, over a total of 1953514552 blocks. Feb 4 01:58:01 Tower ntpd[1189]: synchronized to 72.52.190.26, stratum 2 Feb 4 01:58:01 Tower ntpd[1189]: time reset -0.202804 s Feb 4 01:58:20 Tower ntpd[1189]: synchronized to 72.52.190.26, stratum 2 Feb 4 02:04:36 Tower ntpd[1189]: synchronized to 69.164.197.199, stratum 2 Feb 4 02:07:36 Tower ntpd[1189]: synchronized to 72.52.190.26, stratum 2 Feb 4 02:10:03 Tower ntpd[1189]: synchronized to 69.164.197.199, stratum 2 Feb 4 02:55:15 Tower ntpd[1189]: synchronized to 72.52.190.26, stratum 2 Feb 4 03:00:10 Tower ntpd[1189]: synchronized to 69.164.197.199, stratum 2 Feb 4 09:12:18 Tower kernel: md: sync done. time=26810sec rate=72865K/sec Feb 4 09:12:18 Tower kernel: md: recovery thread sync completion status: 0 syslog-2010-02-04.txt Link to comment
Joe L. Posted February 4, 2010 Share Posted February 4, 2010 Copied 185 Gig to the array and did my first parity check.The first time parity is created it is a parity "sync" (parity is initially written) It is highly recommended you now perform a parity check by pressing the "Check" button on the management interface. at that time parity and all the other data drives will be read and parity re-calculated, and parity compared to what is read from the parity drive. It should not find any parity errors. If it does find errors it usually indicates some kind of hardware problem... Unless you perform this step, there is no way to know you will be able to use the parity data written. To illustrate, let's say a spot on the parity drive was not "written" to an area on the platter than can be read back reliably. You will not know f the manufacturing defect on the sector until you do the parity check. At that point the SMART firmware on the disk will mark the sector as un-readable and mark it as a candidate for re-allocation. When the sector on the disk is NEXT written to it will first attempt to re-write the original location, and if un-successful re-allocate the data in an unused sector from its pool of free sectors. When the SMART firmware determines a sector is un-readable it will return a read error to unRAID. When doing a parity "check" upon a failed read of the parity disk, unRAID will re-write the sector it could not read, fixing the parity to be correct. You really should perform this subsequent parity check just in case there were any bad sectors on the parity disk. Accordingly to the log it took 7h & 45 minutes but accordingly to my watch only 7h & 30 minutes. Well 15 minutes is not the world.I've got no explanation there... Even the time-stamps in the log agree with your watch, but the cumulative seconds sure show 7h45m. Will the parity check time increase as the array gets fuller? It will increase as you add more disks, but will not increase with the existing disks. In fact, it might decrease a tiny bit since the first time you are "writing" the parity disk. All subsequent "checks" will be reading it. Typically read-speeds are faster than write speeds. Now, the speed of the parity disk might not matter at all if it took longer to read all the data disks than write (or read) the parity disk, so as I said, it might not be noticeable. If you follow my advice, you'll soon know if the time for a parity check is the same. Joe L. Link to comment
One2go Posted February 5, 2010 Author Share Posted February 5, 2010 Thanks for the excellent tips. Well it is off and running and doing a parity check. Let's see what will happen 6 hours from now. Link to comment
Joe L. Posted February 5, 2010 Share Posted February 5, 2010 Let's see what will happen 6 hours from now. I think you mean in 7 1/2 to 7 3/4 hours from now... Link to comment
One2go Posted February 5, 2010 Author Share Posted February 5, 2010 Finished and nothing suspicious in the syslog nor on the screen. Below is a sreen shot after the parity check finished. Link to comment
EMKO Posted March 5, 2010 Share Posted March 5, 2010 would 2 Supermicro AOC-SASLP-MV8 work in the 2 pci-e x16 slots? Link to comment
One2go Posted March 8, 2010 Author Share Posted March 8, 2010 I am not sure. I do have an Adaptec 1430SA that is still in original packaging and has not been used because of lack of additional drives. I want to get more of the 2TB Samsung drives (5* on Newegg) but I am not willing to pay a $30 premium per drive from what I paid in December. So once I get additional drives I'll also get this new card which will give me 12 ports on two PCI-e cards and 5 ports from the mobo. This will fill the space for the 17 drives that I have in the Antec 1200. Sorry for not being able to not be more helpful but so far this combination of hardware has performed flawless, with excellent writing speeds both from my i7 950 HP elite and my QNAP NAS TS-239 Pro. Link to comment
One2go Posted March 27, 2010 Author Share Posted March 27, 2010 Well here we go on our way to level 1 testing. Installed the Adaptec 1430SA into the 16x PCI-e slot and connected the new 4 Samsung 2 TB drives to it. Installation went without a hitch. Adaptec card went into configuration by typing Ctrl-A and then in the setup I disabled the BIOS. Rebooted and the 4 drives became visible. Weirdest thing though happened. Since the existing 4 drives were ID /dev/sda to /dev/sdd I expected the new 4 drives to become sde to sdh. But NO the new drives became /dev/sda to /dev/sdd and the old drives are now /dev/sde to /dev/sdh. Man you can wipe out a whole disk of data if you don't read carefully and perhaps blink. When I ran the preclear script I was happy to see the message "Drive does not seem to have a valid partition". Anyhow preclear is on its way and hopefully some 24 hours later these 4 drives come on-line bringing the total to 14 TB. Thanks for all your hard work Joe. Things have been just so easy without any hitch that I am anxiously waiting for the shoe to drop. It just can't be such a breeze with no problems. I do also appreciate the excellent efforts that went into version 4.5, haven't updated to 4.5.3 yet. I really hope that eventually we will have a system with 2 parity disks. Summary so far this mobo has been nothing but fantastic and with the new 8 port controller a total of 21 disks running of PCI-e should be possible. For me adding the 8 port controller will be the end. The mobo has one additional part that I either will use for an additional data disk or should it ever happen a second parity disk. Total of 17 disks max out the Antec 1200. Link to comment
terrastrife Posted March 27, 2010 Share Posted March 27, 2010 arent trying very hard if you can only fit 17 in a 1200 i got space for 20 in a case a bit smaller than the 900 Link to comment
One2go Posted March 28, 2010 Author Share Posted March 28, 2010 I know I could have fit more in the case using 5 in 3 cages but I liked the inexpensive ($ 4 in 3 cages from Scythe ("Hard Disk Stabilizer x4"). No fan and solid, but only 16 drives plus 1 on the bottom. Plus when I planned it Unraid didn't support as many drives as it does now. In addition keep it cool and quiet was an additional desire of mine. With eventually 32 TB of data or 30 TB and 2 parity drives if that ever comes, I am set for awhile. Link to comment
One2go Posted March 29, 2010 Author Share Posted March 29, 2010 The PreClear of the 4 new drives finished and the array is back on line. Just something funny with the PreClear final summary. Here is what showed on the Putty screen: ============================================================================ == == Disk /dev/sda has been successfully precleared == ============================================================================ S.M.A.R.T. error count differences detected after pre-clear note, some 'raw' values may change, but not be an indication of a problem 55c55 < 1 Raw_Read_Error_Rate 0x002f 252 252 051 Pre-fail Always - 0 --- > 1 Raw_Read_Error_Rate 0x002f 100 100 051 Pre-fail Always - 57 67c67 < 195 Hardware_ECC_Recovered 0x003a 252 252 000 Old_age Always --- > 195 Hardware_ECC_Recovered 0x003a 100 100 000 Old_age Always 72c72 < 200 Multi_Zone_Error_Rate 0x002a 252 252 000 Old_age Always - 0 --- > 200 Multi_Zone_Error_Rate 0x002a 100 100 000 Old_age Always - 2 ============================================================================ This summary was identical on all drives with the exception of the very last column. Any idea why that maybe or what this means? Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.