February 26, 201313 yr Author I got v5 RC11 to run on my 128MB flash drive. I made a backup of everything on this drive (a Plus version 4.7). Then I reformated the flash drive. I copied the extracted contents of the RC11 AiO zip file and then copied the config folder from my backup. This lets me run the Basic server version 5 RC11 of unRAID. I'm not going to mount anything with this setup, I just want to use it to preclear the new drive, which I started preclearing last night before going to bed. Once my new 4GB flash drive arrives, I will try to get it working with my Pro license and the correct disk configuration (the configuration I had working a few weeks ago on my 16GB flash drive). The preclear is now at step 2 of 10 (zeroing) and the speed is fairly slow - 10 MB/sec. I am using a different SATA cable, port and controller this time. My motherboard has 9 SATA ports served by 3 different controllers (one controller servers 6 ports SATA1 through SATA6, one serves 2 ports SATA8 and SATA9 and another controller serves 1 port SATA7 plus an eSATA port). I connected the new HD to SATA7 this time. Previously, it was connected to SATA3. I assume this performance ~10 MB/sec is not within the normal range. However, there doesn't seem to be any problem with my hardware. At least nothing that jumps out in the syslog. I think I may try preclearing the new 3TB drive and see if it performs any better. If it preclears at a normal rate, maybe my new 2TB WD20EURS is defective.
February 26, 201313 yr Yes, stick both that 2T and 3T into the box and preclear both. If the 2T is having issues then the 3T should show better performance. I'd also try and stay with the SATA1 to SATA6 ports since those should be the main chipset ports. I'd offer to trade the 3T for a 2T but I doubt it's make much sense with the shipping. I don't have that much luck. I sent back a 1.5T in the fall and got a 1.5T back. Have you gone into the BIOS and turned everything else off that isn't used?
February 26, 201313 yr Author The 3TB drive is currently in the pre-read stage. The performance is considerable faster (128 MB/sec vs. 40 MB/sec) at this stage, but I'm going to wait until the preclear is into the zeroing before drawing any conclusions. I don't know if the performance of the 3TB (WD30EZRX) is normally a lot better than the 2TB (WD20EURS). I didn't want to run both preclears at the same time, fearing one would slow down the other. I was planning to populate my server with 2TB drives by replacing my 1TB drives when it was financially advantageous. I still have 2 - 1TB drives in my array. Then, I was going to start expanding with 4 TB drives if the price came down a significant amount. Whenever, I jump up to the next size, I have to purchase at least 2 drives - 1 for the parity and 1 for data, so in this case I thought I'd wait until the 4TB drives were affordable. Getting a 3TB drive isn't such a bad thing, I think I'll hold onto it and eventually use it in my array as the parity drive for the foreseeable future.
February 27, 201313 yr Author Could someone please comment on what I should expect for the performance during preclearing? The 2TB drive, connected to a couple different ports and using different SATA cables, typically started out at over 50 MB/sec. During the pre-read stage it would gradually slow down. The last time, I think it slowed down to just under 30 MB/sec. Other times, it was maybe a bit slower. Once it was in the zeroing step, it would slow down a lot more. Zeroing would start at 10 to 12 MB/sec, but before even getting to 10%, it would be down below 8 MB/sec. This is slow, but not nearly as slow as the drives I just recently RMA'd, which were 2.5 MB/sec or slower. I'm currently preclearing my new 3TB drive and it started pretty quickly - above 150 MB/sec, during the pre-read stage. However, by the time it completed this part, it too had slowed down a bit, but not nearly as much as the other drive. By the end of the pre-read the 3TB drive was preclearing at around 120 MB/sec. Now it is in the early stage of the zeroing, and it is working at over 150 MB/sec. Again, I'll try not to draw any conclusions because it is only at about 13% of zeroing. What's normal? Based on this, it seems my new 2TB drive might be defective. Does that make sense? This 3TB drive is connected to SATA port5, which I didn't try with the 2TB drive. I will try preclearing the 2TB drive on port5 after the 3TB drive finishes. If the 2TB drive has crappy performance on port5, then maybe I've confirmed it's a bad drive.
February 27, 201313 yr my two 3TB drives took 36 and 39 hours to perform a single pre-clear pass. == invoked as: ./preclear_disk.sh -M 4 /dev/sdg == WDC WD30EZRX-00MMMB0 WD-WCAWZ2982058 == Disk /dev/sdg has been successfully precleared == with a starting sector of 1 == Ran 1 cycle == == Using :Read block size = 8225280 Bytes == Last Cycle's Pre Read Time : 9:09:41 (90 MB/s) == Last Cycle's Zeroing time : 11:13:24 (74 MB/s) == Last Cycle's Post Read Time : 19:19:06 (43 MB/s) == Last Cycle's Total Time : 39:43:13 == == Total Elapsed Time 39:43:13 == == Disk Start Temperature: 26C == == Current Disk Temperature: 36C,
February 28, 201313 yr Author Finally!!! Something worked. The preclear of the 3 TB drive finished about an hour ago. Preclear Successful ... Total time 33:19:25 ... Pre-Read time 7:50:40 (106 MB/s) ... Zeroing time 7:14:04 (115 MB/s) ... Post-Read time 18:13:38 (45 MB/s) So now I'll reconnect the 2TB drive to the same port with the same cable. If it goes very slowly, then I'll return that drive.
February 28, 201313 yr Author The 2 TB drive has been connected using the same SATA cable to the same port as the 3TB drive was and is now at the zeroing step (step 2 of 10). It's only at 4% and the speed is around 10 MB/sec. So the preclear performance of this drive is not due to the hardware it's connected to. Otherwise, the preclear performance of the 3 TB drive would have been similarly slow and it wasn't. Based on past results, I can expect the preclear rate to slow down as it preclears. This could take a week or so. As far as I know, it might stall out. Does anyone know if the performance might be due to the differences between the 2 drives? The 2TB WD20EURS is a SATA3 drive with a maximum transfer rate of 3 GB/sec and the WD30EZRS is a SATA6 drive with a maximum transfer rate of 6 GB/sec. My motherboard (ABIT AB9 Pro) was purchased in 2008, but the date on the user manual is 2006. So it is a 6 - 7 year old design, at least. The manual doesn't say if the SATA ports are 3 GB/s or 6 GB/s, so I assume they were probably just 3 GB/sec. The specifications for newer SATA ports will say 3 or 6. Connected to this motherboard, a 6 GB/sec HD should have no performance advantage over a 3 GB/sec HD, right? Should I assume there is some defect with my new WD20EURS hard drive and apply for an RMA?
February 28, 201313 yr The 2 TB drive has been connected using the same SATA cable to the same port as the 3TB drive was and is now at the zeroing step (step 2 of 10). It's only at 4% and the speed is around 10 MB/sec. So the preclear performance of this drive is not due to the hardware it's connected to. Otherwise, the preclear performance of the 3 TB drive would have been similarly slow and it wasn't. Based on past results, I can expect the preclear rate to slow down as it preclears. This could take a week or so. As far as I know, it might stall out. Does anyone know if the performance might be due to the differences between the 2 drives? The 2TB WD20EURS is a SATA3 drive with a maximum transfer rate of 3 GB/sec and the WD30EZRS is a SATA6 drive with a maximum transfer rate of 6 GB/sec. My motherboard (ABIT AB9 Pro) was purchased in 2008, but the date on the user manual is 2006. So it is a 6 - 7 year old design, at least. The manual doesn't say if the SATA ports are 3 GB/s or 6 GB/s, so I assume they were probably just 3 GB/sec. The specifications for newer SATA ports will say 3 or 6. Connected to this motherboard, a 6 GB/sec HD should have no performance advantage over a 3 GB/sec HD, right? Should I assume there is some defect with my new WD20EURS hard drive and apply for an RMA? i would RMA it. one of the options is "poor performance"
February 28, 201313 yr What he said. Even if the SATA ports were the original SATA I (1.5gbps), that would not make an appreciable difference. As I recall from your previous syslog, the drive was being setup at full speed (3gbps), with no apparent issues reported. I'd say it's the drive, not your hardware. I think it has wasted enough of your time!
March 4, 201313 yr Author The replacement drive arrived today and I've just finished installing and started the pre-clear. I was hoping this replacement would also be a 3TB drive, but that was not to be. They sent another 2TB WD20EURS. Oh well. The performance of the preclear started out at over 90 MB/sec, but has slowed down to around 70 MB/sec and it is at 11% of the pre-read stage.
March 5, 201313 yr Author This is not encouraging. The preclear has slowed to a crawl. At 4% into the zeroing, it has slowed down to under 6MB/sec. At each screen refresh it shows it is getting slower. At this rate it is going to take a very, very long time to preclear this disk. I've shut down my server, disconnected the drive, reconnected it to a different port, rebooted and restarted the preclear. It is just as slow, if not slower. Is there something wrong with these drives? My hardware seems to work ok with the other drives. The preclear of the 3TB WD30EZRS drives performed at a normal speed. Does anyone have any experience preclearing the Western Digital 2TB WD20EURS drives? (Since this thread has gotten so long, most people are probably not reading it anymore, so I'm going to start a new thread specifically about the preclearing question).
March 5, 201313 yr very odd indeed. while it is possible to get two bad replacements its not the norm thats for sure. does this drive have jumpers for 4K aligned or not? are you using the 64 alignment option? what syntax is pre-clear being run with?
March 5, 201313 yr Author I don't see any jumpers for this drive. Here's the syntax I used for the preclear command: ./preclear_disk.sh -A /dev/sdb I chose this drive because the marketing material seemed to indicate that it was better suited for storing audio/video files, which is the primary purpose of my server. It was also priced a little lower than the other 2TB drives at the time and I was happy to save a few dollars. I understood that the primary advantage was that it would be quieter and maybe consume a little less power. I don't really care about the noise because my server is located in a back room in my basement, but conserving power is always nice. I just did a Google search for WD20EURS jumpers and didn't find anything about jumpers, but I did get a thread comparing the WD20EURS to the WD20EARS (http://www.overclock.net/t/876281/wd20ears-vs-wd20eurs). Someone responded that the WD20EURS ignore errors, which might be useful for recording video within a security system. However, this seems like it could be a problem for storing data. I wonder if I can get an RMA and have it replaced with a WD20EARS...
March 5, 201313 yr Author Try booting another system with the USB stick and preclearing the drive on it. I only have one spare desktop PC that I could use to run unRAID and unfortunately, I can't get it to boot to the usb flash drive. In the system bios there are three other boot options - usb-zip, usb-fdd and usb-cdrom and none of them work.
March 5, 201313 yr Author I called Western Digital technical support. After explaining my problem, the tech escalated it to the next level. The next tech speculated there must be an incompatibility with my system and the EURS drives. He's issued me another RMA and will replace the WD20EURS with a WD20EZRX. BTW, I installed the 3TB WD30EZRX and the system bios screen that flashes at startup shows it having a capacity of 801 GB. Once I'm in unRAID, it shows up as a 3TB drive. Is there any reason to be concerned with how my system bios is showing this drive?
March 5, 201313 yr very odd indeed. while it is possible to get two bad replacements its not the norm thats for sure. FYI. Ordered 2 WD 2TB greens from NewEgg. Got them started preclearing one died on pass 1 sent it back to NewEgg. Before I got replacement second one died on pass 2 or 3 so sent that back as well. The replacements were DOA although they did show up in bios just not in unRAID so sent them back separately again. The 3rd round drives were not even recognized by any PC bios I had (there may have been a 4th attempt with NewEgg but I don't remember for sure any more). So I finally just sent them to WD (was tired of the endless loop with NewEgg I appeared to be in) got 3TB greens as upgrade replacements - both new. After several months of use they are now dead/dieing. One of which I mentioned here. These experiences are why I've kept away from WD Green drives lately. Before this incident I had at most a 10% failure rate on the Greens. Had much better luck with WD Red drives. Only 1 drive out of 16 WD Red drives so far have been RMA'd.
March 6, 201313 yr BTW, I installed the 3TB WD30EZRX and the system bios screen that flashes at startup shows it having a capacity of 801 GB. Once I'm in unRAID, it shows up as a 3TB drive. Is there any reason to be concerned with how my system bios is showing this drive? That is how a > 2.2TB drive will look if you don't have a UEFI bios. So booting from it would not be possible but as long as the OS can handle 3TB drives (like unRAID 5.0 and 64 bit Windows OS) you shouldn't have any problems.
March 7, 201313 yr Author I've been using either Hitachi Deskstars or the WD Green drives and haven't had many problems with them until this Fall. The WD Greens had been pretty reliable, especially their 1TB drives. I don't think I ever had any problems with a WD 1TB Green. Anyhow, while waiting for my other warranty replacements, I ran preclear on the 3TB drive again. This time connected to the port (SATA3) that it will use after it is in the correct physical slot. In this case I moved it into the 3rd slot to replace disk2. It completed the 2nd preclear in 34 hours and 16 minutes. However, after running a SMART status report it shows there is 1 Current_Pending_Sector. 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 - 17 3 Spin_Up_Time 0x0027 196 184 021 Pre-fail Always - 5175 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 9 5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0 7 Seek_Error_Rate 0x002e 100 253 000 Old_age Always - 0 9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 77 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 - 9 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 6 193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 20 194 Temperature_Celsius 0x0022 128 109 000 Old_age Always - 22 196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 1 198 Offline_Uncorrectable 0x0030 100 253 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 100 253 000 Old_age Offline - 0 SMART Error Log Version: 1 There were 0 pending sectors after the 1st preclear, which I think was run from a different port (SATA5 or SATA7). I'm going to try running preclear again. If it doesn't show 0 Current_Pending_Sectors after this preclear, what should I do?
March 7, 201313 yr If it doesn't show 0 Current_Pending_Sectors after this preclear, what should I do? You could keep running preclears until it does go to zero or you can RMA it as a bad drive because of the pending sector. A sector that is pending means that you can't depend on it to have good data and that could affect a rebuild of another disk.
March 8, 201313 yr Author All of my replacement drives have been delivered. The replacement for the WD20EURS, a WD20EZRX and a Hitachi Deskstar 5K2000 drive finally arrived. So I intstalled all of these drives and using screen, started preclearing all of them. If the 3T WD30EZRX doesn't show 0 Current_Pending_Sectors after its 3rd preclear, I will get an RMA and send it back. I will also check the SMART status reports of the WD20EZRX and the Hitachi drive to make sure they are perfect before adding to my array. The preclears have been running since yesterday and all seemed to be running at a normal pace. They've slowed down a bit since then. If you monitor the preclear performance with unMenu, the performance is slower. According to the status displayed in the unMenu MyMain screen, the 3TB WD30EZRX is preclearing (still zeroing) at a rate of 27 MB/s, the 2TB drives (WD20EZRX and Hitachi Deskstar 5K2000) are now in the post-read stage at a rate of 38 MB/s. However, if you check the status on the command console and toggle to the different screens, the 3TB WD30EZRX is preclearing at 27 MB/s (the same as shown in unMenu) and the 2TB drives are preclearing at 83.8 MB/s (Hitachi 5K2000) and 75.2 MB/s (WD20EZRX). Why is there a difference between the unMenu results and the command console, especially for the 2TB drives? A much bigger issue/problem are the errors showing up in my syslog. What do these errors mean? Do I have a bad drive, controller, SATA controller, motherboard...? Here's a peak at the last 15 lines: Mar 8 10:54:33 Media kernel: res 40/00:b4:90:1c:83/00:00:b7:00:00/40 Emask 0x10 (ATA bus error) (Errors) Mar 8 10:54:33 Media kernel: ata5.00: status: { DRDY } (Drive related) Mar 8 10:54:33 Media kernel: ata5.00: failed command: WRITE FPDMA QUEUED (Minor Issues) Mar 8 10:54:33 Media kernel: ata5.00: cmd 61/00:a0:90:24:83/04:00:b7:00:00/40 tag 20 ncq 524288 out (Drive related) Mar 8 10:54:33 Media kernel: res 40/00:b4:90:1c:83/00:00:b7:00:00/40 Emask 0x10 (ATA bus error) (Errors) Mar 8 10:54:33 Media kernel: ata5.00: status: { DRDY } (Drive related) Mar 8 10:54:33 Media kernel: ata5.00: failed command: WRITE FPDMA QUEUED (Minor Issues) Mar 8 10:54:33 Media kernel: ata5.00: cmd 61/00:a8:90:20:83/04:00:b7:00:00/40 tag 21 ncq 524288 out (Drive related) Mar 8 10:54:33 Media kernel: res 40/00:b4:90:1c:83/00:00:b7:00:00/40 Emask 0x10 (ATA bus error) (Errors) Mar 8 10:54:33 Media kernel: ata5.00: status: { DRDY } (Drive related) Mar 8 10:54:33 Media kernel: ata5.00: failed command: WRITE FPDMA QUEUED (Minor Issues) Mar 8 10:54:33 Media kernel: ata5.00: cmd 61/00:b0:90:1c:83/04:00:b7:00:00/40 tag 22 ncq 524288 out (Drive related) Mar 8 10:54:33 Media kernel: res 40/00:b4:90:1c:83/00:00:b7:00:00/40 Emask 0x10 (ATA bus error) (Errors) Mar 8 10:54:33 Media kernel: ata5.00: status: { DRDY } (Drive related) Mar 8 10:54:33 Media kernel: ata5: hard resetting link (Minor Issues) Every time I try to attach my syslog (20MB zipped), this message doesn't get posted and the "new topic" screen appears. Ok, it didn't upload because even zipped, the syslog is too big. Here's a link to the file: http://we.tl/NUO88QyOTr
March 8, 201313 yr Author After a quick search of "ATA bus errors" I found this info here: http://lime-technology.com/wiki/index.php/The_Analysis_of_Drive_Issues#Drive_Interface_Issues I think the closest case is probably "Drive interface issue #4" which says the problem is usually related to a backplane connection, which I don't have. Otherwise, this suggests that I probably have a cable connection problem, which is weird because I made sure the connections were very solid after installing the hard drives and booting up the server. I also have a brand new power supply, but that's no guarantee that it is working right. Is there any way to determine if the problem is with the power cable or the SATA cable? If the problem is with the power cable, it might be a challenge to diagnose. Like most power supplies (mine is a Corsair CX500), it has several SATA power connections and some 4-pin molex connectors. If you run out of SATA connectors, you can attach a 4-pin molex to SATA adapter cable. In the case of SATA05 on my system, the drive is connected to one of the straight SATA power connections, and not one that uses the adapter cable. One other question... Is a 500Watt power supply sufficient for a server with 8 or 9 drives? Should I have purchased a 600W or 650W instead?
March 8, 201313 yr Author Here's a clip from an earlier section of my syslog: Mar 8 10:46:50 Media kernel: Descriptor sense data with sense descriptors (in hex): Mar 8 10:46:50 Media kernel: 72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00 Mar 8 10:46:50 Media kernel: b6 f0 1f b0 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] ASC=0x0 ASCQ=0x0 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] CDB: cdb[0]=0x2a: 2a 00 b6 ef 77 b0 00 04 00 00 Mar 8 10:46:50 Media kernel: end_request: I/O error, dev sde, sector 3069147056 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] Result: hostbyte=0x00 driverbyte=0x08 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] Sense Key : 0xb [current] [descriptor] Mar 8 10:46:50 Media kernel: Descriptor sense data with sense descriptors (in hex): Mar 8 10:46:50 Media kernel: 72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00 Mar 8 10:46:50 Media kernel: b6 f0 1f b0 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] ASC=0x0 ASCQ=0x0 Mar 8 10:46:50 Media kernel: sd 4:0:0:0: [sde] CDB: cdb[0]=0x2a: 2a 00 b6 ef af b0 00 04 00 00 Mar 8 10:46:50 Media kernel: end_request: I/O error, dev sde, sector 3069161392 Mar 8 10:46:50 Media kernel: ata5: EH complete Mar 8 10:46:50 Media kernel: ata5.00: exception Emask 0x10 SAct 0x7fffffff SErr 0x400100 action 0x6 frozen Mar 8 10:46:50 Media kernel: ata5.00: irq_stat 0x08000000, interface fatal error Mar 8 10:46:50 Media kernel: ata5: SError: { UnrecovData Handshk } Mar 8 10:46:50 Media kernel: ata5.00: failed command: WRITE FPDMA QUEUED Mar 8 10:46:50 Media kernel: ata5.00: cmd 61/00:00:b0:57:f0/04:00:b6:00:00/40 tag 0 ncq 524288 out Mar 8 10:46:50 Media kernel: res 40/00:14:b0:5f:f0/00:00:b6:00:00/40 Emask 0x10 (ATA bus error) Mar 8 10:46:50 Media kernel: ata5.00: status: { DRDY } Mar 8 10:46:50 Media kernel: ata5.00: failed command: WRITE FPDMA QUEUED Mar 8 10:46:50 Media kernel: ata5.00: cmd 61/00:08:b0:5b:f0/04:00:b6:00:00/40 tag 1 ncq 524288 out Mar 8 10:46:50 Media kernel: res 40/00:14:b0:5f:f0/00:00:b6:00:00/40 Emask 0x10 (ATA bus error) Mar 8 10:46:50 Media kernel: ata5.00: status: { DRDY } This shows what appear to be a problem with the "sde" drive. This is the 3TB WD30EZRX and it is connected to the SATA3 connection on my motherboard. Does "I/O error, dev sde, sector 3069147056" mean it is having problems writing to that sector and will result in a 'Current_Pending_Sector' added to my count? I'm not sure what ata5.00 is. I'm assuming it means the SATA5 connection, which would be the connection for the new Hitachi drive. Or, maybe it is for Port05, which would be the connection for one of my older 1 TB drives that is currently my disk5 on the SATA6 connection.
Archived
This topic is now archived and is closed to further replies.