Pauven Posted July 23, 2012 Share Posted July 23, 2012 I have 6 WD Red 3TB drives arriving today, and will spend the next several days pre-clearing them. I'll post back later on my impressions of the drives. I'm expecting it will take me a week to integrate them into the array. I currently have 18 Samsung drives (F1's, F2's and F3's), some of which are 3.5 years old, most of which are over 2 years old, and none of which have ever failed. I have at least 4 other Samsung F1's in various computers in the home, and none of those have ever failed. Never had a DOA either. I know everyone's experience is unique, and not all have been as fortunate as I have been, but my word that is an amazing track record. I briefly had a 2TB Western Digital RE4-GP as a parity drive, a really expensive high-end 5-year warranty bugger, but it died within a couple months. I was so P/O'd I never even had it replaced (but with that 5 year warranty, I guess it's not too late). I had no intention of abandoning the Samsung drives, and a very high reluctance to go back to Western Digital, but in the current HD market the new Red's just seemed like a logical choice. Really hoping I don't get burned again. Quote Link to comment
Pauven Posted July 25, 2012 Share Posted July 25, 2012 Finished a 1 pass pre-clear on 3 of the WD Red 3TB drives with flying colors (everything that matters is still a zero). Of the remaining 3 drives, 1 may be defective (didn't show on boot) and 2 are still in the wrapper. Here's the pre-clear reports: Jul 25 04:03:30 Tower preclear_disk-diff[26320]: ========================================================================1.13 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == invoked as: ./preclear_disk.sh -A /dev/sdv Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == WDC WD30EFRX-68AX9N0 WD-WMC1T0077429 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Disk /dev/sdv has been successfully precleared Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == with a starting sector of 1 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Ran 1 cycle Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Using :Read block size = 8225280 Bytes Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Last Cycle's Pre Read Time : 7:55:20 (105 MB/s) Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Last Cycle's Zeroing time : 7:12:13 (115 MB/s) Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Last Cycle's Post Read Time : 18:04:48 (46 MB/s) Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Last Cycle's Total Time : 33:13:22 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Total Elapsed Time 33:13:22 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Disk Start Temperature: 33C Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Current Disk Temperature: 30C, Jul 25 04:03:30 Tower preclear_disk-diff[26320]: == Jul 25 04:03:30 Tower preclear_disk-diff[26320]: ============================================================================ Jul 25 04:03:30 Tower preclear_disk-diff[26320]: ** Changed attributes in files: /tmp/smart_start_sdv /tmp/smart_finish_sdv Jul 25 04:03:30 Tower preclear_disk-diff[26320]: ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Jul 25 04:03:30 Tower preclear_disk-diff[26320]: Temperature_Celsius = 120 117 0 ok 30 Jul 25 04:03:30 Tower preclear_disk-diff[26320]: No SMART attributes are FAILING_NOW Jul 25 04:58:14 Tower preclear_disk-diff[6222]: ========================================================================1.13 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == invoked as: ./preclear_disk.sh -A /dev/sds Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == WDC WD30EFRX-68AX9N0 WD-WMC1T0076840 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Disk /dev/sds has been successfully precleared Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == with a starting sector of 1 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Ran 1 cycle Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Using :Read block size = 8225280 Bytes Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Last Cycle's Pre Read Time : 8:09:59 (102 MB/s) Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Last Cycle's Zeroing time : 7:26:46 (111 MB/s) Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Last Cycle's Post Read Time : 18:29:42 (45 MB/s) Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Last Cycle's Total Time : 34:07:29 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Total Elapsed Time 34:07:29 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Disk Start Temperature: 42C Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Current Disk Temperature: 34C, Jul 25 04:58:14 Tower preclear_disk-diff[6222]: == Jul 25 04:58:14 Tower preclear_disk-diff[6222]: ============================================================================ Jul 25 04:58:14 Tower preclear_disk-diff[6222]: ** Changed attributes in files: /tmp/smart_start_sds /tmp/smart_finish_sds Jul 25 04:58:14 Tower preclear_disk-diff[6222]: ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Jul 25 04:58:14 Tower preclear_disk-diff[6222]: Temperature_Celsius = 116 108 0 ok 34 Jul 25 04:58:14 Tower preclear_disk-diff[6222]: No SMART attributes are FAILING_NOW Jul 25 06:35:12 Tower preclear_disk-diff[23441]: ========================================================================1.13 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == invoked as: ./preclear_disk.sh -A /dev/sdt Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == WDC WD30EFRX-68AX9N0 WD-WMC1T0073984 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Disk /dev/sdt has been successfully precleared Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == with a starting sector of 1 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Ran 1 cycle Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Using :Read block size = 8225280 Bytes Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Last Cycle's Pre Read Time : 8:23:43 (99 MB/s) Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Last Cycle's Zeroing time : 7:40:33 (108 MB/s) Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Last Cycle's Post Read Time : 19:01:48 (43 MB/s) Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Last Cycle's Total Time : 35:07:05 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Total Elapsed Time 35:07:05 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Disk Start Temperature: 41C Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Current Disk Temperature: 33C, Jul 25 06:35:12 Tower preclear_disk-diff[23441]: == Jul 25 06:35:12 Tower preclear_disk-diff[23441]: ============================================================================ Jul 25 06:35:12 Tower preclear_disk-diff[23441]: ** Changed attributes in files: /tmp/smart_start_sdt /tmp/smart_finish_sdt Jul 25 06:35:12 Tower preclear_disk-diff[23441]: ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Jul 25 06:35:12 Tower preclear_disk-diff[23441]: Temperature_Celsius = 117 109 0 ok 33 Jul 25 06:35:12 Tower preclear_disk-diff[23441]: No SMART attributes are FAILING_NOW Quote Link to comment
Helmonder Posted July 26, 2012 Share Posted July 26, 2012 Just started preclear on a 3TB RED :-) Quote Link to comment
Pauven Posted July 29, 2012 Share Posted July 29, 2012 Update: Out of 6 WD Red 3TB drives, 1 was DOA, and 1 appears to have died during clearing. Of the 4 that survived, one is now Parity, one replaced a 1TB data drive, and the other 2 are pending data drive upgrades. So, for me a 33% failure rate, as compared to my 0% failure rate for my Samsungs. Man I miss Samsung. I did have a power-failure, apparently long enough to drain my UPS, near the end of the preclear on the last two drives, and the power outage appears to have been between hour 35 and 36 of the Preclear, so I'm thinking the Preclear actually finished before the power outage. The drive that survived has a valid Preclear signature, while the other drive is the one that is now dead. I'm thinking I will run a another preclear on the drive the lived, even though the SMART report is flawless. Helmonder, how did your preclear go? Quote Link to comment
WeeboTech Posted July 30, 2012 Share Posted July 30, 2012 while the other drive is the one that is now dead. on the drive that is dead, is it totally doa? undetected? or bad sectors? I know if my machine powers off abnormally. a drive can end up with bad sectors. Quote Link to comment
Dephcon Posted July 30, 2012 Share Posted July 30, 2012 Rebuilding parity onto my Red at 98MB/s, I'm satisfied with the performance of these drives. Quote Link to comment
Pauven Posted July 30, 2012 Share Posted July 30, 2012 Both drives that are dead exhibit the same issue: They don't show in BIOS, and during startup unRAID sees them and tries to connect to them. These connection attempts go on for several minutes, delaying the server startup. Eventually it looks like Linux gives up on them and unRAID finishes booting. Inside the GUI, the dead drives are not visible. So these drives are so dead I can't get a SMART report. Here's the last set of error messages for the two drives. Earlier error messages have the same errors (softreset failed / SRST failed type messages). Jul 29 16:56:28 Tower kernel: ata19: softreset failed (device not ready) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata20: softreset failed (device not ready) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata20: limiting SATA link speed to 1.5 Gbps (Drive related) Jul 29 16:56:28 Tower kernel: ata20: softreset failed (device not ready) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata20: reset failed, giving up (Minor Issues) Jul 29 16:56:28 Tower kernel: ata19: softreset failed (device not ready) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata19: softreset failed (1st FIS failed) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata19: limiting SATA link speed to 1.5 Gbps (Drive related) Jul 29 16:56:28 Tower kernel: ata19: softreset failed (device not ready) (Minor Issues) Jul 29 16:56:28 Tower kernel: ata19: reset failed, giving up (Minor Issues) And yes, before you ask, I have tried these drives on other slots / controllers in the server. The only thing I haven't done with them yet is plug them into my Windows machine to see how they behave on another computer. Quote Link to comment
Helmonder Posted July 30, 2012 Share Posted July 30, 2012 Update: Out of 6 WD Red 3TB drives, 1 was DOA, and 1 appears to have died during clearing. Of the 4 that survived, one is now Parity, one replaced a 1TB data drive, and the other 2 are pending data drive upgrades. So, for me a 33% failure rate, as compared to my 0% failure rate for my Samsungs. Man I miss Samsung. I did have a power-failure, apparently long enough to drain my UPS, near the end of the preclear on the last two drives, and the power outage appears to have been between hour 35 and 36 of the Preclear, so I'm thinking the Preclear actually finished before the power outage. The drive that survived has a valid Preclear signature, while the other drive is the one that is now dead. I'm thinking I will run a another preclear on the drive the lived, even though the SMART report is flawless. Helmonder, how did your preclear go? I did three cycles and flying colours, it is in the array rebuilding now. Quote Link to comment
chickensoup Posted July 31, 2012 Share Posted July 31, 2012 So out of 7 drives that have been purchased in this thread we have 1 DOA, correct? I haven't included the second drive that @Pauven reported died as it was under dirty circumstances (power failure during preclear). Even if the preclear likely finished it's hard to say whether the drive was at fault here. I've never blamed a drive for being at fault when there's an environmental issue involved, you can't really- unless the UPS forced a clean shutdown. Still, unfortunate to hear that @Pauven lost two drives Quote Link to comment
Johnm Posted August 4, 2012 Share Posted August 4, 2012 We picked up 6 of the 3TB RED's for a NAS at work. We have a 6 drive NAS at work that has 6 Samsung F4's (2 TB each) in RAID6 configuration. One of our samsungs F4's finally died after 2 years of 24x7 use. instead of replacing it, we decided to expand the NAS from a 6 drive RAID6 NAS from 8TB to 12TB 1 DOA and 1 filled with smart errors. the 4 good ones we configured as a RAID5 and stress tested them.... the TYLER seems to work. (we could not use the WD greens in it at all) and they did as advertised. Once we got the 2 RMA drives back, we rebuilt the array as a 6 drive RAID6 and all was well. the NAS itself reported the RED drives as slower then the Samsung F4's they replaced.. but the array still exceeded the speed of the copper Gigabit (That is all we care about for our needs). they are pretty quiet and seem to do the job as expected. So far we are happy with the results. I might do the same thing myself at home.. If I see the 3TB RED's at a reasonable price, I will upgrade one of my 4drive Samsung F4 RAID5's with 3 TB RED's and move the F4's to my secondary unraid. the 33% fail rate is bad... but for all we know it is not WD's drives fault, It could have happened in shipping. I will point out that we overpaid for them. We had to use CDW for a vendor and they were about $225 each with shipping and tax for the 3TB flavor when we ordered ours. Since then, the price has gone down. Quote Link to comment
downloadski Posted August 9, 2012 Share Posted August 9, 2012 What will unraid do when a WD Red drive reports back that it cannot read the sector anymore. (TLRE will not try it much and report dead sector i think) It seem the file is corrupted for Unraid, will Unraid re-create the missing data on a other part of the disc somehow ? It seems a good option when you run a raid setup which should correct a broken sector by remapping the data. If the OS or controller does not do this, you have corrupted files.. Quote Link to comment
Helmonder Posted August 9, 2012 Share Posted August 9, 2012 I guess unraid will throw a red ball and take the array offline. Drive should be replaced after which parity will recreate the data. Quote Link to comment
downloadski Posted August 9, 2012 Share Posted August 9, 2012 So not the drives for me than.. Have 4 TB Hitachi's but the price of them seem to increase. The 3 TB Hitachis are hard to find and more expensive than the WD red, so i thought it was an option with the 3 year warrantee. But red balled drives i do not want. Quote Link to comment
Helmonder Posted August 9, 2012 Share Posted August 9, 2012 Well.... This is just the way it works with every drive.. Difference with the WD RED's seems to be that the drive will try for a shorter amount of time to fix the issue itself. This means you will now earlier that something might be going wrong.. In the end that means less chance of dataloss.. You could have a drive that would give no indication of failure untill it goes up in flames.. Would you prefer that ? I prefer to know as soon as possible that a drive might be starting to fail.. It will make it possible for me to lessen the chance of loosing data, also makes it possible to give a drive like that an extremely rigoreous pre-clear to make it fail even sooner after the first signs, that makes it more likely that I will be able to fall back on guarantee (the three years itself helps also to that effect) Quote Link to comment
downloadski Posted August 9, 2012 Share Posted August 9, 2012 Will WD not ask you to run the test tool and if it passes that it is not accepted for RMA ? Quote Link to comment
Helmonder Posted August 9, 2012 Share Posted August 9, 2012 Sure they will... But if a drive starts to show failure then a rigorous preclear will do either of two things: - drive will further deteriorate up to the point it could be RMA'd; - failure rate remains stable and you can just reuse the drive. Everybody happy.. Quote Link to comment
bonzi Posted August 9, 2012 Share Posted August 9, 2012 I got 3x 3TB reds a few weeks ago from NCIX when they first came out. I precleared them, no errors. They've been sitting in my array for the past three or so weeks. Quote Link to comment
downloadski Posted August 12, 2012 Share Posted August 12, 2012 Helmonder, good points. I have a offer for secondhand Hitachi 7200 RPM 2 GB HDD's of less than a year old for 80 euro. Tempting.. Quote Link to comment
Helmonder Posted August 12, 2012 Share Posted August 12, 2012 I would not go for 7200's... They run more hot and with parity on the added speed is not that impressive (disk is not the bottleneck) .. But to each his own ! Btw: 2TB's are a bit over 100 eur over here, I would go for new ones, that saves you the guarantee.. Pricedifference is soon gone if one fails in half a year.. Quote Link to comment
downloadski Posted August 12, 2012 Share Posted August 12, 2012 I gues looking at your name 'here' is the same for me http://tweakers.net/pricewatch/cat/333/serial-ata-harde-schijven.html#filter:q1ZKKkrMS_FMKVayilYyUorVUUrLLCouCSjKTE4NycxN9U2sULLKK83JwZDIzINJ5BelpBa5ZabmpChZKZVlppYXK-koFYAUghUZm-ooFSfnF0HMgnNAUgZATmpOanJJakpwQWoy1BnmQO0mZiZKsbUA 105 Euro it seems. That migh be wiser indeed, but at a shop i bought something else, they could only deliver 2 and told me the brand will be phased out. It will be all WD i assume than. Quote Link to comment
Helmonder Posted August 12, 2012 Share Posted August 12, 2012 I have allmost all WD 2TB Green, they perform fine .. Quote Link to comment
carlos28355 Posted August 31, 2012 Share Posted August 31, 2012 just read through this thread. So it seems to that these drives will be good for unraid by finding out sooner rather than (too) later that the drive is failing? everyone still happy with them? i will probably make my next hdd a red but im not rushing out to swap my drives out for no reason! thanks for the info! Quote Link to comment
Helmonder Posted August 31, 2012 Share Posted August 31, 2012 So far, so good: 1 Raw Read Error Rate 0x002f 200 200 051 Pre-fail Always Never 0 3 Spin Up Time 0x0027 179 177 021 Pre-fail Always Never 6050 4 Start Stop Count 0x0032 100 100 000 Old age Always Never 138 5 Reallocated Sector Ct 0x0033 200 200 140 Pre-fail Always Never 0 7 Seek Error Rate 0x002e 200 200 000 Old age Always Never 0 9 Power On Hours 0x0032 099 099 000 Old age Always Never 853 10 Spin Retry Count 0x0032 100 100 000 Old age Always Never 0 11 Calibration Retry Count 0x0032 100 253 000 Old age Always Never 0 12 Power Cycle Count 0x0032 100 100 000 Old age Always Never 11 192 Power-Off Retract Count 0x0032 200 200 000 Old age Always Never 1 193 Load Cycle Count 0x0032 200 200 000 Old age Always Never 137 194 Temperature Celsius 0x0022 124 106 000 Old age Always Never 26 196 Reallocated Event Count 0x0032 200 200 000 Old age Always Never 0 197 Current Pending Sector 0x0032 200 200 000 Old age Always Never 0 198 Offline Uncorrectable 0x0030 100 253 000 Old age Offline Never 0 199 UDMA CRC Error Count 0x0032 200 200 000 Old age Always Never 0 200 Multi Zone Error Rate 0x0008 100 253 000 Old age Offline Never 0 Quote Link to comment
dikkiedirk Posted August 31, 2012 Share Posted August 31, 2012 Every drive failing is a PITA, but will eventually happen with any brand. Even more so when out of warranty. WD (4 died last year) has been less painless, with their advance replacement. Samsung is also quit easy going through Seagate but no advance replacement. Nowadays the warranty period has decreased unfortunately. Are the WD Reds the only ones with 3 year warranty? Any Dutch or European sources for WD Reds? Quote Link to comment
StevenD Posted September 4, 2012 Share Posted September 4, 2012 Has anybody run across these recently, in-stock and not $300?? I purchased an ARC-1220 and I want two of the 3TB REDs to use as my parity/cache. Thanks! 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.