1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

HP ProLiant MicroServer Owners Club! (Attempting to sell here will result in bans)

Discussion in 'Storage & Backup' started by oli, May 10, 2011.

  1. wasas

    wasas Member

    Joined:
    Jun 14, 2002
    Messages:
    720
    Location:
    /au/qld/bne
    mmm My current 6x1tb's in raid are fine, along with another 3x 1tb's in my media.. and a 500gb torrent drive that has been smashed for the last 36 months. They are all going strong still.

    The reason I asked was around the head parking, and issues it may cause. as they will run 24/7 and no doubt always be spinning I dont think it'll be a issue at all.

    http://www.umart.com.au/pro/products_listnew.phtml?id=10&bid=2&id2=129
    The prices for all the 2tb's in the low end are similar, so some seagates or WD's would be my pick, unless someone can prove they are slower or inferior.
     
  2. stumo

    stumo Member

    Joined:
    May 17, 2011
    Messages:
    85
    Location:
    VIC
  3. OP
    OP
    oli

    oli Member

    Joined:
    Jun 29, 2002
    Messages:
    7,263
    Location:
    The Internet
  4. Blinky

    Blinky Member

    Joined:
    Jul 4, 2001
    Messages:
    4,181
    Location:
    Brisbane
  5. Kapz786

    Kapz786 Member

    Joined:
    May 22, 2011
    Messages:
    5
    Location:
    Perth, WA
    Anyone got recommendations on a 4gig stick of ECC Ram they are using with this set up?

    Or 2gig x 2 ?
     
  6. Stanza

    Stanza Member

    Joined:
    Jun 27, 2001
    Messages:
    2,901
    Location:
    Adelaide
  7. mikehol

    mikehol Member

    Joined:
    Jun 15, 2006
    Messages:
    58
    Yeah, thats what I thought. "synonymous with Hitachi" ha! It was definitely in the IBM days that the title was earned. And IBM generally had a very good reputation for hardware quality. Maybe that's why they offloaded their hard drives to Hitachi. Anyway, all brands are much the same now.
    And they are all so cheap, there is no excuse not to have RAID (raid1 software will do fine) if your data has value and backups are not enough.
     
  8. OP
    OP
    oli

    oli Member

    Joined:
    Jun 29, 2002
    Messages:
    7,263
    Location:
    The Internet
    Scorptec is on the ball with this as well. I placed my order on the 17th. Got a phone call on the 18th I think and was told they expected to get them in on the 24th and would be posting them out on the 24th and/or 25th. :thumbup:

    I checked tonight and their site says 'to be dispatched' so I guess I'll have it on Thursday if it's done via AAE or whatever other courier service isn't retarded... :p

    Agreed. I just buy whatever is good value at the time and if it dies I have redundancy with RAID and if that dies I restore from backups.
     
  9. SaTaN

    SaTaN Member

    Joined:
    Jun 18, 2002
    Messages:
    4,790
    Location:
    Caulfield-ish
    not that one but yes, I'm putting a 20char*4 line display in once I write some software for it :p (lcdproc development doesnt seem to be active still)
     
  10. antenna

    antenna Member

    Joined:
    Feb 26, 2007
    Messages:
    537
    Location:
    SA
    I realise this probably sounds ridiculously stupid but I got an internal converter thing (well ok, there's 2 things :weirdo:) for the ODD bay so I can use a 3.5 HDD there, problem is I can't at all figure out how the things are meant to be fitted in there or where the screws they supply are meant to go, or what that pushy lever thing does. Yeah so.. totally confused, any help appreciated. :o
     
  11. Oblong Cheese

    Oblong Cheese Member

    Joined:
    Aug 31, 2001
    Messages:
    10,595
    Location:
    Brisbane
    Took me a while as well. The "ODD" screws from inside the front door go into the bottom rung of screw holes on your 5.25" item (CDROM, drive caddy, whatever) then it simply slides into the slot and locks into place.
     
  12. GooSE

    GooSE New Member

    Joined:
    Jun 26, 2001
    Messages:
    6,679
    Location:
    Sydney
    Thanks a lot, that was an extremely informative post. It makes me worry a lot less about hard drive choice and configuration now.

    BTW, did you use gnop -S 4096 or a modified version of zpool?

    Changing the topic a little... It would be great if we could somehow gain control of the blue/red HP light on the front. I believe it's called the health light or something.
     
  13. antenna

    antenna Member

    Joined:
    Feb 26, 2007
    Messages:
    537
    Location:
    SA

    Thanks :thumbup:

    It sounds so easy and yet I'm still lost, lol. I can't even figure out how the rail things I got work with the HDD I have. :facepalm:
     
  14. Oblong Cheese

    Oblong Cheese Member

    Joined:
    Aug 31, 2001
    Messages:
    10,595
    Location:
    Brisbane
  15. Hyram

    Hyram Member

    Joined:
    Jan 19, 2009
    Messages:
    827
    Geez, get out of the last century, will ya!

    There were only three Deathstar drives from IBM, and one of the very first things Hitachi did when they took over was figure out what went wrong and fix it. For much of the last decade, Hitachi drives were considered the best of the best, and they carried a price tag to match. As ol' Vincent Price used to tell us, "At Hitachi there's no room for compromise." About 18 months ago or so, they finally succumbed, closed down most of their production lines in Japan, and copied everyone else by having their gear made in the shitpits of Shenzen, China. There was a noticable (but not very big) drop in quality, but Hitachi went for the jugular and put RRPs on their drives that gave Sadgate & WD the willies, and Hitachi's enterprise-grade UltraStars make WD Caviar Blacks look like a box of crayons.

    Hitachi are also conservative, they never adopted a new technology until everyone else finished cutting their teeth with it. They were the last to adopt perpendicular recording, for example.

    If you want low noise, low power and high reliability, there's really only one choice, Hitachi's CoolSpins.
     
  16. Hyram

    Hyram Member

    Joined:
    Jan 19, 2009
    Messages:
    827
    If I remember to find a camera the next time I go to the office, I'll take a snap of the RMA drive-stacks. WD Greens outnumber everything else at least three to one.
     
  17. Hyram

    Hyram Member

    Joined:
    Jan 19, 2009
    Messages:
    827
    You don't mention the brand or model or anything else useful, howzabout a link to what you bought and another to where you bought from?
     
  18. rugger

    rugger Member

    Joined:
    Aug 24, 2003
    Messages:
    813
    Location:
    Perth, WA
    Linux mdadm performance

    Finally got some iozone performance numbers for my mdadm raid 5 array, consisting of 4 old WD Green 1TB drives.

    Server is stock, standard 1gig ram, slackware linux installed. CFQ scheduler for disks, formatting correctly in ext4 using 256kb chunk size. Write cache in bios is enabled.

    Only tuning I've done is

    echo 4096 > /sys/block/md0/md/stripe_cache_size

    Results are:

    Code:
    	Iozone: Performance Test of File I/O
    	        Version $Revision: 3.397 $
    		Compiled for 64 bit mode.
    		Build: linux-AMD64 
    
    	Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
    	             Al Slater, Scott Rhine, Mike Wisner, Ken Goss
    	             Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
    	             Randy Dunlap, Mark Montague, Dan Million, Gavin Brebner,
    	             Jean-Marc Zucconi, Jeff Blomberg, Benny Halevy, Dave Boone,
    	             Erik Habbinga, Kris Strecker, Walter Wong, Joshua Root,
    	             Fabrice Bacchella, Zhenghua Xue, Qin Li, Darren Sawyer.
    	             Ben England.
    
    	Run began: Tue May 24 00:36:08 2011
    
    
    	Machine = Linux gamma 2.6.33.4 #3 SMP Wed May 12 23:13:09 CDT 2010 x86_64 AMD 	Include fsync in write timing
    	Include fsync in write timing
    	CPU utilization Resolution = 0.001 seconds.
    	CPU utilization Excel chart enabled
    	Record Size 128 KB
    	File size set to 16000000 KB
    	Percent read in mix test is 70
    	Command line used: ./iozone -M -e -+u -T -t 1 -r 128k -s 16000000 -i 0 -i 1 -i 2 -i 8 -+p 70 -C
    	Output is in Kbytes/sec
    	Time Resolution = 0.000001 seconds.
    	Processor cache size set to 1024 Kbytes.
    	Processor cache line size set to 32 bytes.
    	File stride size set to 17 * record size.
    	Throughput test with 1 thread
    	Each thread writes a 16000000 Kbyte file in 128 Kbyte records
    
    	Children see throughput for  1 initial writers 	=  128030.25 KB/sec
    	Parent sees throughput for  1 initial writers 	=  127991.42 KB/sec
    	Min throughput per thread 			=  128030.25 KB/sec 
    	Max throughput per thread 			=  128030.25 KB/sec
    	Avg throughput per thread 			=  128030.25 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU Utilization: Wall time  124.970    CPU time   26.628    CPU utilization  21.31 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =  128030.25 KB/sec, wall=124.970, cpu=26.628, %= 21.31
    
    	Children see throughput for  1 rewriters 	=  128455.37 KB/sec
    	Parent sees throughput for  1 rewriters 	=  128443.05 KB/sec
    	Min throughput per thread 			=  128455.37 KB/sec 
    	Max throughput per thread 			=  128455.37 KB/sec
    	Avg throughput per thread 			=  128455.37 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time  124.557    CPU time   21.134    CPU utilization  16.97 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =  128455.37 KB/sec, wall=124.557, cpu=21.134, %= 16.97
    
    	Children see throughput for  1 readers 		=  140217.98 KB/sec
    	Parent sees throughput for  1 readers 		=  140216.21 KB/sec
    	Min throughput per thread 			=  140217.98 KB/sec 
    	Max throughput per thread 			=  140217.98 KB/sec
    	Avg throughput per thread 			=  140217.98 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time  114.108    CPU time   13.071    CPU utilization  11.45 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =  140217.98 KB/sec, wall=114.108, cpu=13.071, %= 11.45
    
    	Children see throughput for 1 re-readers 	=  142011.53 KB/sec
    	Parent sees throughput for 1 re-readers 	=  142011.02 KB/sec
    	Min throughput per thread 			=  142011.53 KB/sec 
    	Max throughput per thread 			=  142011.53 KB/sec
    	Avg throughput per thread 			=  142011.53 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time  112.667    CPU time   13.140    CPU utilization  11.66 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =  142011.53 KB/sec, wall=112.667, cpu=13.140, %= 11.66
    
    	Children see throughput for 1 random readers 	=   10608.37 KB/sec
    	Parent sees throughput for 1 random readers 	=   10608.36 KB/sec
    	Min throughput per thread 			=   10608.37 KB/sec 
    	Max throughput per thread 			=   10608.37 KB/sec
    	Avg throughput per thread 			=   10608.37 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time 1508.243    CPU time   25.558    CPU utilization   1.69 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =   10608.37 KB/sec, wall=1508.243, cpu=25.558, %=  1.69
    
    	Children see throughput for 1 mixed workload 	=   13327.79 KB/sec
    	Parent sees throughput for 1 mixed workload 	=   13327.51 KB/sec
    	Min throughput per thread 			=   13327.79 KB/sec 
    	Max throughput per thread 			=   13327.79 KB/sec
    	Avg throughput per thread 			=   13327.79 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time 1200.499    CPU time   21.212    CPU utilization   1.77 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =   13327.79 KB/sec, wall=1200.499, cpu=21.212, %=  1.77
    
    	Children see throughput for 1 random writers 	=   12848.67 KB/sec
    	Parent sees throughput for 1 random writers 	=   12848.46 KB/sec
    	Min throughput per thread 			=   12848.67 KB/sec 
    	Max throughput per thread 			=   12848.67 KB/sec
    	Avg throughput per thread 			=   12848.67 KB/sec
    	Min xfer 					= 16000000.00 KB
    	CPU utilization: Wall time 1245.265    CPU time   22.050    CPU utilization   1.77 %
    
    	Child[0] xfer count = 16000000.00 KB, Throughput =   12848.67 KB/sec, wall=1245.265, cpu=22.050, %=  1.77
    
    
    
    iozone test complete.
    
    One thing I've noticed is that my random write performance is very much lower then the ZFS numbers. I suspect fsync() isn't working correctly on that system, or the system is rewriting the random writes into a sequential write stream like an SSD does. As ZFS is a transactional filesystem (with images and stuff), I'm actually going to suspect that it the later ... converting the random writes into linear writes.
     
  19. Imitation

    Imitation Member

    Joined:
    Jun 27, 2001
    Messages:
    2,868
    picoPSU is the brand and he said earlier he got the 80w model.

    http://www.mini-box.com/picoPSU-80
     
  20. Stanza

    Stanza Member

    Joined:
    Jun 27, 2001
    Messages:
    2,901
    Location:
    Adelaide
    Post a pic or link to the rails you bought antenna, we will help you get it mounted :thumbup:
     

Share This Page

Advertisement: