Kaygee Posted September 29, 2009 Share Posted September 29, 2009 Parity check completed, memtest 24hr burn-in done. Data copied to array md5 checked. Pretty good speed for parity check too. Disk share copy 35MB/s. Sep 29 22:31:46 Tower kernel: md: sync done. time=14080sec rate=69372K/sec Sep 29 22:31:46 Tower kernel: md: recovery thread sync completion status: 0 Quote Link to comment
kapperz Posted September 29, 2009 Share Posted September 29, 2009 I think I'm the first to use a GIGABYTE GA-MA785G-UD3H motherboard. What is needed to level 1 test? How many level tests are needed to add a motherboard to the compatibility list? So far its been working with rosewill raid cards, but I have not assigned a parity drive yet. I'm still in the process of copying my data. Quote Link to comment
prostuff1 Posted September 29, 2009 Share Posted September 29, 2009 check out the Motherboard Rating System thread on a discussion. Also take a look on the Hardware Compatibility Page and look through the boards that have been tested, they will lead you to the specific threads that you can base your testing on. Quote Link to comment
Kaygee Posted September 30, 2009 Author Share Posted September 30, 2009 After looking at the syslog it seems the Seagate drives in the system dont like AHCI mode. They dissappeared during the parity check. I switched to Native IDE and am re-running the parity check. I did try legacy IDE but the syslog showed all drives as UDMA100 so I skipped that mode. Data copy checked for speed (disk share write 35MB/s), MD5 checked OK. Parity check seems to be running at a more consistent speed (85-95MB/s), whereas in AHCI it would bounce between 70-90MB/s, finishing the parity check at 69MB/s. I'll post the syslog once the check is completed. Quote Link to comment
Kaygee Posted September 30, 2009 Author Share Posted September 30, 2009 Syslog from last nights parity check. Native IDE mode selected. Seagate drives are still seen by the system. Parity check speed was slightly better also which is a bonus. Force NCQ disabled = yes. 76MB/s was the final figure upon completion. Summary AMD 780G NB and SB700 SB. ATI Radeon 3200 onboard video. Socket AM2+ uATX form factor. 6 x sata ports (Legacy SATA, Native SATA, RAID and AHCI) 1 x Gb Marvell NIC (88E8056). No e-Sata ports. Additional info Supports 8GB ram (4GB installed). Four slots, DDR2 1066 (requires AM2+ CPU). 1 x PCI-e x16 slot 1 x PCI-e x1 slot 2 x PCI 10 x USB (EHCI compliant) (four on backplane, six onboard). Supports MPS 1.1/1.4 Environmental CPU temp 39 C/35 C, MB temp 35 C, HDD temp idle 24 C, load 35 C. CPU info AMD 3250e low power dual core 1500. ADJ3250IAA5DO. AMD dual core TDP 22W (11w per core). Quote Link to comment
Kaygee Posted October 1, 2009 Author Share Posted October 1, 2009 Expanded the array by adding another drive, added a cache drive too. Parity generaton triggered. Oct 1 00:53:07 Tower kernel: md: sync done. time=25330sec rate=38561K/sec Oct 1 00:53:07 Tower kernel: md: recovery thread sync completion status: 0 Rather pleased with the results. 4 x 1TB, Adding 2 x 500GB for Parity creation test. Parity creation Parity check Disk share write Load testing Adv load test Previous pci system 14MB/s 25MB/s 25MB/s 5 streams 5 streams no writes Pci-e system 38MB/s 76MB/s 35MBs 5 streams 5 streams with writes Performance was acceptable previously, however writing to an in use (streamng) disk could cause some stuttering. Parity generation and checking are @ three times faster, disks share writes also increased noticably. Being able to write whilst reading is a big and unexpected plus. No more lsof for me. This is also with a low power cpu installed, admitedly dual core vs single core. Disk share writes are based on file sizes transferred, bwm-ng shows Gb throughput averages around 45MB/s. System pci pci-e differences cpu 3500+ 2 x 1500 cpu type Sempron Athlon x2 memory ddr 400 ddr2 800 sata controller sil pci sb700(amd) * vga fx5200 ati radeon 3200 Gb rt8169 yukon * native ide mode ncq disabled = yes Quote Link to comment
Kaygee Posted October 11, 2009 Author Share Posted October 11, 2009 OK managed to break it, added another two drives and it no longer boots the USB device. Seems max drives bios can handle is seven and still boot the usb device. Tried latest bios, beta bios, switched usb device to force fdd as per some asus boards, the onlt thing I havnt tried is adding a floppy (dont have one!). Revisiting the XFX Geforce 8300 board to see if we can get that to work. Quote Link to comment
prostuff1 Posted October 11, 2009 Share Posted October 11, 2009 You might look at the Topical Index and the PloP Boot Manager section. It might allow you to boot unRAID a slightly different way and still allow for more drives. Quote Link to comment
Kaygee Posted October 11, 2009 Author Share Posted October 11, 2009 Yes, I thought of PLoP (I used it very successfully on my original unRAID server - excellent little tool) , but I dont run a cache disk, prefering to copy direct to disk shares. If both of the other two available motherboards wont work, I may revisit the Abit A-S78H but for now it is being re-tasked as a VMWARE testbed server. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.