BobPhoenix

Drives vary. I have some that are slow as well including a Seagate drive that can only do 35MB/s read and write access as a stand alone drive. Nothing shows on smart reports for it since it came with my N40L I didn't think I would be able to return it. If you feel it is too slow you can return it to WD as a performance problem - at least the last time I returned a drive to WD that was still available anyway.

You guys are correct. Must have worked for me with 4.7 but it doesn't with 5.0rc16c. Added the "force user = root" to my initial post.

I didn't use the vmdk's in this thread. I created my own in VMWare and formatted it in a Windows VM as FAT32 maybe that is why it is working as hda1 for me. As to writing can you write to it using MC on unRAID directly? If you can then it is in your smb-extra.conf file as that is where the permissions for writing from the network are controlled. If you can't write to if from unRAID itself then it is the mount that needs to be changed in the GO file. The -w on the mount command should make it writable. I started out on this from unMENU. I mounted it with unMENU and looked at the top of the screen to see the mount command used then copied it and changed the name of the mount point to be one I wanted and put it in my GO file and tested writes with MC. Then I looked on the forum for smb-extra.conf directions and experimented until I got it working for write across the network.

How do you know what drive to mount? I tried with /dev/hda1 and the share just shows empty. When I do /dev/sda1 it seems be the flash drive. You have to change that /dev/hda1 to the vmdk if yours is not the first IDE vmdk on your VM then you might need /dev/hdb1 for instance. I only have a single IDE device defined in the VM hence the /dev/hda1. You need the 1 so that it points at the partition not the drive for the mount point. I got that part of the code from unMENU which is what you can also use if you only want to make this temporarily available. unMENU will mount it and share it for you. I just put it in my go file and smb-extra.conf to make it ALWAYS available. I use IDE on my VMs for the vmdk because I've never been able to make unRAID boot off of a SATA vmdk. To use unMENU to temporarily mount the drive just go to the Disk Management tab and scroll down to the bottom you will see something like attachment. The buttons to mount and share are at the bottom. You have to mount it twice. The first time just mounts it read only then gives you the option to remount read/write. Then share it out. Next boot it will be gone or if you unshare and unmount it. Edit: Also by making my vmdk IDE I don't have to worry about unRAID/Linux changing the boot device location since it is the ONLY IDE HDD on the VM. Edit2: Must admit I didn't try too hard to find the problem with booting to a SATA vmdk since I had this in mind when I setup my first ESXi build.

I just have this in my GO file after emhttp to map the vmdk to a mount point: mkdir /mnt/boothdd mount -w -t vfat /dev/hda1 /mnt/boothdd 2>&1 I also share it with an smb-extra.conf: [boothdd] comment = Public Folder force user = root path = /mnt/boothdd public = yes writable = yes Then I just copy across the network to both the flash share and my boothdd share and reboot the unRAID VM - done. The only time I have to do anything is when I need to run make_bootable.bat again. Edit: fixed smb-extra.conf to allow writing correctly with 5.0rc16c thanks to RockDawg.

I thought I was getting faster results to cache drive then that on my N40L but I'm at work currently and can't test for you. I will see when I get home tonight. Maybe someone else can confirm earlier. Writing to a user share which writes to the cache drive I get 50-60MB/s. Writing directly to the cache drive share and not through the user share file system I get 75-85MB/s. This was using Windows Explorer in a Windows 7 VM from one unRAID server to the N40L unRAID server. I copied a BluRay folder of a movie (34.8GB) - "2001 A Space Odyssey". The cache drive is a EADS 1.5TB green drive and the rest of the drives in the array are WD 2TB Reds.

I thought I was getting faster results to cache drive then that on my N40L but I'm at work currently and can't test for you. I will see when I get home tonight. Maybe someone else can confirm earlier.

I've never tried it on another Linux distro but I suspect it will not work.

You can use the free version of unRAID to pre-clear drives. I have two boxes I regularly pre-clear drives on. I've done up to 5 drives at a time and they were all from the free version of unRAID. When not pre-clearing drives they are Windows Desktop PCs.

I agree. Plus you would have to remove any pass through cards in order to take a snapshot - at least that is true for me at ESXi 5.0 anyway.

Yes I knew that I think the speed has improved slightly. In the VM setting, it was set to: Number of virtual sockets: 1 Number of cores per sockets: 2 I have changed other way round to: Number of virtual sockets: 2 Number of cores per sockets: 1 I don't understand what that mean but the transfer speed seem to be improved, what did you set to? This: Number of virtual sockets: 1 Number of cores per sockets: 2 simulates a multi core CPU. This: Number of virtual sockets: 2 Number of cores per sockets: 1 simulates a multi CPU MB Or at least that is what I am assuming. I can't say I've seen anything to confirm that theory.

Never used the non-RAID version only have the RR622 and I've only used them with Windows not Linux. That might be why they work better for me. The Marvel drivers for Windows may be better then the Sil drivers and the opposite for Linux. For me speed wasn't an issue as I use them JBOD to Windows and not in any RAID configuration where simultaneous access is required. So I've never bench marked them in Windows simultaneous or separately. I routinely get 80-100+MB/s copying files 4GBs+ between drives but that is only two drives at a time.

Thanks for pointing out the need for a low-profile card. I overlooked that (I don't have one of these systems) ... and I'm surprised Sans-Digital included cards on its list that didn't meet that criteria. As for the 622 not supporting port multipliers .... where did you see that note? I looked at the Newegg link and it's not on any of their descriptive pages, and in fact one of the reviewers noted: "... Wanted a card that supports port multiplication to just access individual drives. It wasn't clear in the specs, at the maufacturer's website, or in the reviews that this card will do anything besides RAID or JBOD, but it works fine ..." That, plus the fact it's on Sans Digital's list of "eSATA Port Multiplier" cards that they reference in their product descriptions would tend to support that it should work! [ http://www.sansdigital.com/esata-port-multiplier/index.php ] Have you actually tried that card? I have not tried the card, I saw the comment, I cannot remember where it is. What I can say for sure, the silicon image cards work with port multiplers. The Marvel chipsets work, but not as smoothly as the silicon image chipsests. The Asmedia chipsets work, but again, not as smoothly as the silicon image chipsets. In my tests I would do 1 dd, then 2 in parallel, then 3 in parallel, then 4. With the silicon image chipsets, Each successive parallel dd lowered the throughput by 30MB/s. I.E. one test was almost 120MB/s, two were at 60MB/s, Three around 40MB/s, 4 at 30MB/s. These were my findings, YMMV. By the end of the test, 4 drives were DD'ing at 30MB/s each at the same time with a very smooth access of data to each drive. With the other chipsets access was not as smooth. With the ASmedia Chipset, it worked by by the time I had 4 in parallel, the access was on the order of 10MB/s for each drive. Access was very blocky. I.E. When one process was DD'ing. It seemed to lock out (block) access to the other drives. The Marvel wasn't as fast as the Silicon image chipset. (from my findings years ago, current technology may be better). I've been playing with these port multipliers ever since linux started supporting them, these are my findings from testing. There may be better cards. I can only afford so many controllers to test with. (grin). I've used a SYBA Silicon Image 3124 chipset on PCIe x1, Addonics Silicon Image 3132 on PCIe x1 and the ASmedia SATA III PCIe x1 card on this particular box. With the Startech ASmedia Card, SATA III and an SSD, I was able to achieve over 350MB/s on PCIe x1. With a 3TB SATA III 7200 RPM drive I was able to achieve over 190MB/s. Port multiplier speed suffered greatly when accessing multiple drives simultaneously. Before any expenditure, I would suggest testing the eSATA port on the back of the box to see how well it performs. The high point rocket card may work and perform well with the port multipliers. However, there still is the low profile point that has to be considered. Someone else in the thread was exploring a 4 port card. Perhaps that card could be considered. The model of 622 that you need is the RocketRAID 622. That one DOES support port multiplying. I'm currently using two they work real well. I've got a couple of Sil3132s and they were dropping drives continually. With the RocketRAIDs I have very few drops unlike the SIL3132s. Not sure why my SILs were so bad because I read some of the same things about that SILs are better then Marvel but for me it has been the opposite. Haven't tried the ASmedia controllers yet.

But can I do that with the N54L? I've been using external enclosures for over two years with my first unRAID. So I understand what speeds I will get. During the parity check I get speeds around 30 MB/s with twenty data drives. I use a cache drive so I can get good speeds when transferring data and the cache drive transfers it to the array later. I just want to make sure this is possible with the N54L. My other unRAIDs I setup with regular motherboards. But the N54L is a micro PC. Basically I corrected myself and was saying the speed difference might not be noticable and maybe as little as I see with a SAS expander. Just didn't want to delete my post so I corrected it.

Using port multipliers will slow down the speed of a parity check or rebuild - possibly quite a bit. Accessing the data should be fine however. Actually after thinking about it a bit it might be as little as a SAS Expander does - so may not even notice it.

I would do a little more research to see if your board supports VT-D and if ESX will handle the Pass through. I don't know enough to guide you, but I'm not sure the i3 will support it either. What we know works is the popular Supermicro X9SCM boards and the Sandy Bridge XEON CPU along with the IBM 1015 controller. I believe there is a hack for the super micro controller. Again. I'm not advanced on the topic. I would suggest reading JohnM's atlas thread. Supermicro works, and its not really a "hack", its just a specific piece of configuration you need to do for a lot of passthru devices.. Is there a source of this information for 'other' devices? Link for the SASLP-MV8, http://lime-technology.com/forum/index.php?topic=7914.msg128847#msg128847 however I'm interested in where to obtain information on other controllers that may support pass through with this sort of information. I've not had to do the hack for anything but the MV8 - if I remember correctly. I've got AVerMedia Duet tuners, Hauppauge HVR-2250 tuners, USB 2.0 cards (PCI and PCIe), HighPoint 1742 & 622 HDD controllers and IBM M1015 HDD controllers all on passthrough. Only the MV8 has ultimately required the hack to the vmx files for the VMs. I have put the HDD controllers in the "passthru.map" file if ESXi doesn't recognize them. All the information to put into the "passthru.map" file can be obtained in vCenter Client for ESXi - namely the VENDOR ID and the DEVICE ID - if you don't want to use "lspci" from the command line (Like Me). I always set "fptShareable" to false so that the device in only used in one VM and not by any other VM or ESXi itself. Better explanation is here: http://www.vmware.com/pdf/vsp_4_vmdirectpath_host.pdf

Got one of those covers for where the optical drive opening is. It looks excellent but the bracket also is a very tight fit, tighter then the two drives with Nexus Double Twin adapter, but it works. Yes. I tried to make mine bend out in the middle of the bracket and split the difference a little bit inside the case at the outside edges and a little bulge sticking out in the middle. Wish they could reduce it in width by a couple of millimeters so it would fit a little better.

I'm using this with the double twin in my N40L. It is a tight fit in mine but I got it to work.

If your solution doesn't work and if windows can do a long format then you could try that but I don't think you will get it to format in windows if your solution doesn't clear it. You could also try badblocks or SpinRite and see if they can recover the drive but I doubt it is a disk surface problem (what they are good for) and more an electronics/cabling/power problem like Joe L said. I'm betting that even though it was working in windows before either the drive is defective now or you have a cable/power problem.

I had a drive like that. As a last attempt before I RMA'd it I took the drive to a Windows PC and did a long format. After that completed with no errors I was able to run 3 preclear cycles on it like normal. SO long story short you could try writing to the drive first by using the '-n' parameter of preclear for a single cycle first. Then run however many preclear cycles you wanted to run. That is essentially what I did with a windows long format.

I agree with Joe, that combo of drive model and firmware is not confidence inspiring. Any chance of a firmware update for it? Probably - I've got some 2TB greens with 81.xxx I believe. At this point unless it is out of waranty I'm not going to mess with a firmware update - just RMA the drive. I have 2 other 3TB models - they may all be the same model - will have to check when I get home. Only one of the 3 is working correctly. The other problem drive is slow and has some of the same smart errors as this one just not the pending. At one point I really liked the green drives - have 30 currently. But have 12 with smart errors I don't trust - multi zone, read errors etc... another 12 were perfectly fine until I fried them with a case fan that died and they got to ~60c on a parity check. Of the 6 2 TB drives left in working condition at least one is slow now - 30 MB/s parity checks when in array. I finally got the drives replaced with Hitachi/WD Reds in 2TB and 3TB sizes in my 2 unRAID servers. I could RMA most of these but I no longer have a good idea of which ones were over heated and which ones weren't. I don't think they should pay for my mistake anyway and I can use them for offline backups until they are dead.

FYI. The drive I posted about has been in a box to RMA back to WD several times and now several months (just haven't had the opportunity to start the process yet). I got it out last night and put it back in my preclear station to get the smart report - like I've done several times before. The 65535 is a permanent reading until I initiate another preclear cycle. It also isn't always a 65535, 0, 65535. The first set of 3 preclear cycles I ran was 65535 cycle 1, 128? cycle 2 and 0 cycle 3 - so I thought it was a good drive. The last time I ran a cycle of 3 (there may have been other cycles in-between) it was 65535, 0, 65535. So I haven't run one since so that it is in that state when I send it back to WD. So far every time I put it back into my preclear station(s) (I have two cpus that I clear drives on) it comes back with the same smart report I posted above.

I've got a 3TB WD Green drive that flip/flops pending sectors. One pass it's 65535 then next it's zero then 65535 then zero etc.. for at least 6 passes. I'm going to RMA since the last pass left it at 65535. That was the highest I've ever seen too. It was a drive WD upgraded me too when I RMA'd another defective 2TB drive to them. It was the "last straw" for me on WD green drives just had too many go bad on me lately. Only had 1 out of 10 WD Red's bad so far which is much better than the 30% failure rate of Seagate drives and 40% of WD Greens. Sure wish Hitachi and Samsung still made drives. Guess I'll have to try some Toshiba besides the WD Red's. Oh they were all purchased in 1's and 2's mostly from newegg but not all and as I said even straight from WD RMA process. And for me I know it isn't a power supply as it was the only HDD in a preclear station that has cleared 90% of all of the drives I've got some before and after this one. It's just a bad drive . smartctl 5.40 2010-10-16 r3189 [i486-slackware-linux-gnu] (local build) Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net === START OF INFORMATION SECTION === Device Model: WDC WD30EZRS-11J99B1 Serial Number: WD-WCAWZ0773504 Firmware Version: 80.00A80 User Capacity: 3,000,592,982,016 bytes Device is: Not in smartctl database [for details use: -P showall] ATA Version is: 8 ATA Standard is: Exact ATA specification draft version not indicated Local Time is: Wed Feb 27 19:11:21 2013 CST SMART support is: Available - device has SMART capability. SMART support is: Enabled === START OF READ SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED General SMART Values: Offline data collection status: (0x84) Offline data collection activity was suspended by an interrupting command from host. Auto Offline Data Collection: Enabled. Self-test execution status: ( 0) The previous self-test routine completed without error or no self-test has ever been run. Total time to complete Offline data collection: (52980) seconds. Offline data collection capabilities: (0x7b) SMART execute Offline immediate. Auto Offline data collection on/off support. Suspend Offline collection upon new command. Offline surface scan supported. Self-test supported. Conveyance Self-test supported. Selective Self-test supported. SMART capabilities: (0x0003) Saves SMART data before entering power-saving mode. Supports SMART auto save timer. Error logging capability: (0x01) Error logging supported. General Purpose Logging supported. Short self-test routine recommended polling time: ( 2) minutes. Extended self-test routine recommended polling time: ( 255) minutes. Conveyance self-test routine recommended polling time: ( 5) minutes. SCT capabilities: (0x3035) SCT Status supported. SCT Feature Control supported. SCT Data Table supported. SMART Attributes Data Structure revision number: 16 Vendor Specific SMART Attributes with Thresholds: ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE 1 Raw_Read_Error_Rate 0x002f 200 200 051 Pre-fail Always - 21 3 Spin_Up_Time 0x0027 161 147 021 Pre-fail Always - 8950 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 52 5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0 7 Seek_Error_Rate 0x002e 200 200 000 Old_age Always - 0 9 Power_On_Hours 0x0032 095 095 000 Old_age Always - 4028 10 Spin_Retry_Count 0x0032 100 253 000 Old_age Always - 0 11 Calibration_Retry_Count 0x0032 100 253 000 Old_age Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 46 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 41 193 Load_Cycle_Count 0x0032 199 199 000 Old_age Always - 5400 194 Temperature_Celsius 0x0022 130 104 000 Old_age Always - 22 196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0 197 Current_Pending_Sector 0x0032 066 066 000 Old_age Always - 65535 198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 4 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 405 200 Multi_Zone_Error_Rate 0x0008 200 200 000 Old_age Offline - 3 SMART Error Log Version: 1 No Errors Logged SMART Self-test log structure revision number 1 No self-tests have been logged. [To run self-tests, use: smartctl -t] SMART Selective self-test log data structure revision number 1 SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS 1 0 0 Not_testing 2 0 0 Not_testing 3 0 0 Not_testing 4 0 0 Not_testing 5 0 0 Not_testing Selective self-test flags (0x0): After scanning selected spans, do NOT read-scan remainder of disk. If Selective self-test is pending on power-up, resume after 0 minute delay.

I found a utility for dos, but now I got this errot Failed to initialize PAL Any clue what that means ?? //Peter That sounds similar to flashing a M1015 on an incompatible MB.

Is your data important or easily replaceable? Also if you don't use preclear then when you add a disk to the array the whole array will be unavailable while unRAID writes zeros to the drive. The time will vary based on drive size but for a 2TB drive it will likely be ~9 hours. If you preclear it first your array stays available while the drive is being precleared. The other reason to preclear a drive is to exercise a drive to find a bad drives. Most drives die within the first ~100 hours or run to their warranty period or longer. Thanks man for the details, I'm new here and learning all the time. So I guess I'll stop being impatient and just preclear all my drives This is going to take 1 - 2 weeks to have up and running with plugins shares per clears etc but slow and steady wins the race I guess. I was the same way when I started. I did run preclear because I read about the array being off line until unRAID cleared it so I ran a single cycle. Ended up loosing a drive when it died and I had to replace it so then I went to what someone else recomended - 3 cycles. Thought about going to more but saw another discussion about diminishing returns so settled on 3 cycles as MY best compromise. Sure you can run more cycles and probably catch more faults but you also put more ware on the drive that way too. Now I preclear all drives even those that are not headed for the unRAID system with 3 cycles.