June 6, 201511 yr ... I wish there was a way to directly monitor the persistent cache, it would be really interesting to see it in action. Definitely agree. Of course Seagate hasn't released any details on this at all ... what we do know is due to some re-engineering work by 3rd party testing organizations who disassembled the drive; added a window to the outer casing; re-assembled it (all in a clean room of course); then took videos of the drive in action. But it'd be really neat to have a graphic that showed the current state of the persistent cache and the number of pending band rewrites Don't, however, hold your breath ... I doubt Seagate is going to release any nifty little driver that will provide that level of detail (if their firmware even supports providing it externally ... which I suspect it does, as surely they do some internal testing to "tune" the performance).
June 6, 201511 yr ... serves NO clients at all so in normal operation the 3 drives are spun down all day until 12am (when I am in bed anyway) when the daily incremental backup kicks in from the onboard W10 VM. Nifty way to do it ... let Windows do the backup without the need for a separate PC Was there any issue installing Win 10 in a KVM VM ?
June 6, 201511 yr ... serves NO clients at all so in normal operation the 3 drives are spun down all day until 12am (when I am in bed anyway) when the daily incremental backup kicks in from the onboard W10 VM. Nifty way to do it ... let Windows do the backup without the need for a separate PC Was there any issue installing Win 10 in a KVM VM ? A couple, but none that couldn't be intuitively resolved. 1) W10 Setup didn't find any HDD to install to. So I downloaded the latest virtue iso here: https://fedorapeople.org/groups/virt/virtio-win/direct-downloads/archive-virtio/virtio-win-0.1.105-1/ and installed the amd64 viostor for W8.1 driver at the "Load Driver" prompt in W10 Setup. Once loaded Setup found the Virtual Disk I created in VM Setup in Unraid. 2) On setup restart the machine "hung" and went into paused mode. When I tried to resume I got an error from Unraid KVM. I had to destroy the VM from the command line using commands I found here: http://www.cyberciti.biz/faq/linux-kvm-stop-start-guest-virtual-machine/ 3) Windows did not have a built in network card driver. So I used the amd64 NetKVM for W8.1 driver on the above iso and it found the network device and installed fine. There are a few "other" PCI Devices which W10 can't find a driver for but I haven't looked at them yet. They are not interfering with the operation of the VM. P.S. Back to topic, I am getting a bit carried away with the Seagate 8TB Sticky post. You might have to reign me in when you review the post!
June 7, 201511 yr ... P.S. Back to topic, I am getting a bit carried away with the Seagate 8TB Sticky post. You might have to reign me in when you review the post! Good to know you're working on it -- I was about ready to remind you I'll take a look at it when you post it ... I'm sure it will be fine. Just try to summarize rather than providing excessive amounts of details/charts => it's much more useful that way. You can always include links to some of the more detailed posts in this thread for those who want to see all of the raw data. On the other hand, you DO want enough detail that it's clear these have been thoroughly tested (you've done a REALLY good job of that). By the way, these little guys are on sale for $240 through tomorrow here: http://lime-technology.com/forum/index.php?topic=40441.msg380827#msg380827 (Doesn't help you guys in Aussieland, but for those in the US ... ) ... I'm not ready to build my next server, but I did order one just for kicks [i'll stick it in my workstation.] I'm actually thinking or ordering 3 more tomorrow just to toss together a little 3-drive UnRAID system to "play" with the 8TB drives ... but I REALLY don't need that right now !!
June 8, 201511 yr ... My guess is that the noise level is low enough that when the drives are mounted in a case, the normal noise level from the fans probably masks this sound ... Correct. I agree that the "seek" sound of these drives is a bit louder than others, but not loud enough to raise any worries. Also, it's exemplar dependent - two out of my three 8TB Seagate Archive drives are "ticking" noticeably louder, one - same or almost same as other brands and models. Maybe Seagate forgot to put sound insulation in.
June 8, 201511 yr ... Maybe Seagate forgot to put sound insulation in. Oil. As someone already said, they probably forgot to put some oil...
June 8, 201511 yr ... Maybe Seagate forgot to put sound insulation in. Oil. As someone already said, they probably forgot to put some oil... If it was oil missing, it would be creaking, not ticking...
June 8, 201511 yr Just got my server setup with four of these drives. I want to start preclearing. Is there a faster preclear script other than the default and are there any specific settings recommended for these drives?
June 8, 201511 yr Brian's version does the post-read in about half the time of Joe's script: http://lime-technology.com/forum/index.php?topic=32564.0 I've not used it, but from the reports it seems to be just fine. Note that if you're creating a new server (i.e. not adding these to an existing server) the pre-clear doesn't save any time when adding the drives ... so you're just using it to test the drive integrity. For this purpose you could also just use the manufacturer's diagnostics (SeaTools) or even another manufacturer's tests [i use WD's Data Lifeguard to test all of my drives, regardless of manufacturer].
June 8, 201511 yr Brian's version does the post-read in about half the time of Joe's script: http://lime-technology.com/forum/index.php?topic=32564.0 I've not used it, but from the reports it seems to be just fine. Note that if you're creating a new server (i.e. not adding these to an existing server) the pre-clear doesn't save any time when adding the drives ... so you're just using it to test the drive integrity. For this purpose you could also just use the manufacturer's diagnostics (SeaTools) or even another manufacturer's tests [i use WD's Data Lifeguard to test all of my drives, regardless of manufacturer]. So you're saying instead of doing a pre-clear I could just boot to a bootable version of Data Lifeguard and test my drives that way before adding them to the array (which I realize will take a good deal of time itself)? What do you recommend? A single extended test per drive? You can't run them in conjunction can you?
June 8, 201511 yr No, you can't run multiple tests at once with Data Lifeguard. I generally test new drives like this: (a) Run a Quick Test (b) Run an Extended Test © Do a full write zeroes (d) Repeat (a) and (b) If ANY errors I return the drive for a replacement. Otherwise I put it in service. If I have multiple drives to test at once, I just attach them to different systems [i have several ] Note that SeaTools allows multiple drives to be tested at once, so that's another alternative. ... or you can just use the pre-clear script. ALL of these will take a long time with 8TB drives.
June 10, 201511 yr A fairly easy way I check my drives is with an external dock connected to my laptop via eSATA cable. I don't have to physically install the drive or handle it, since it just plugs into the dock easily. It is a basic Windows 7 laptop that has a program called HD TUNE on it. This can do a sector by sector check test, stress test and other tests to your new drive. When satisfied just pull it from the dock and install in your unRaid box. Of course if you don't have a separate desktop/laptop you can't do this. The dock is something like this: http://www.newegg.com/Product/Product.aspx?Item=N82E16817153066 And I think HD TUNE gives you 30 days trial period too. http://www.hdtune.com/
June 11, 201511 yr Ok ran my first preclear on all 4 drives. Took about 57 hours. What am I looking for in these preclear reports with regard to errors/flags? preclear_rpt_Z8405DRY_2015-06-11.txt preclear_rpt_Z8405E0F_2015-06-11.txt preclear_rpt_Z84051A6_2015-06-11.txt preclear_rpt_Z84052S8_2015-06-11.txt
June 11, 201511 yr Ok ran my first preclear on all 4 drives. Took about 57 hours. What am I looking for in these preclear reports with regard to errors/flags? All of these drives are perfect scores. Only perfection is acceptable, but sometimes that can be achieved with additional preclears. No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 1. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 1. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. ============================================================================
June 11, 201511 yr Ok ran my first preclear on all 4 drives. Took about 57 hours. What am I looking for in these preclear reports with regard to errors/flags? All of these drives are perfect scores. Only perfection is acceptable, but sometimes that can be achieved with additional preclears. No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 1. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 1. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. ============================================================================ That's what I figured but the smart data always confuses me. But thanks for confirming this for me. 2 more preclears to go!
June 11, 201511 yr In addition to the preclear_rpt file you should look at preclear_finish => this shows the final SMART report data. If nothing significant has changed and there are no reallocated sectors or pending reallocations (as shown in the summary you posted from preclear_rpt) all is PROBABLY well; but it's a good idea to confirm that none of the less significant parameters have any significantly degraded values. With WD drives, that's fairly easy to tell from the raw report; with Seagate drives there are a few parameters they report that will almost always show less-than-perfect values but are nothing to worry about ... the key is whether or not the "Value" column has degraded (and even then you can ignore High_Fly-Writes). If you post the preclear_finish data I can tell you if I see anything to be concerned about.
June 11, 201511 yr In addition to the preclear_rpt file you should look at preclear_finish => this shows the final SMART report data. If nothing significant has changed and there are no reallocated sectors or pending reallocations (as shown in the summary you posted from preclear_rpt) all is PROBABLY well; but it's a good idea to confirm that none of the less significant parameters have any significantly degraded values. With WD drives, that's fairly easy to tell from the raw report; with Seagate drives there are a few parameters they report that will almost always show less-than-perfect values but are nothing to worry about ... the key is whether or not the "Value" column has degraded (and even then you can ignore High_Fly-Writes). If you post the preclear_finish data I can tell you if I see anything to be concerned about. Thanks gary. preclear_finish_Z8405DRY_2015-06-11.txt preclear_finish_Z8405E0F_2015-06-11.txt preclear_finish_Z84051A6_2015-06-11.txt preclear_finish_Z84052S8_2015-06-11.txt
June 11, 201511 yr They all look fine. Seagate reports the raw read errors ... WD does not. ALL drives have errors when reading the platters ... that's what the ECC data is for (so they can be corrected). You'll note that the raw read # always matches the "ECC recovered" #. Since both of these are typically big numbers, it's easy to wonder if something's wrong ... but as long as both are the same, everything's normal. WD only reports any errors that couldn't be corrected (i.e. actual errors). Seagate also reports raw seek error rates (again, these are normal) ... WD only reports actual errors.
June 12, 201511 yr They all look fine. Seagate reports the raw read errors ... WD does not. ALL drives have errors when reading the platters ... that's what the ECC data is for (so they can be corrected). You'll note that the raw read # always matches the "ECC recovered" #. Since both of these are typically big numbers, it's easy to wonder if something's wrong ... but as long as both are the same, everything's normal. WD only reports any errors that couldn't be corrected (i.e. actual errors). Seagate also reports raw seek error rates (again, these are normal) ... WD only reports actual errors. Good to know. Thanks for confirming this.
June 20, 201511 yr Since my testing of these drives I have bought a couple more. I know in my testing I did three cycles executed separately and not using the -C switch but for those who want to know how long it takes to do a 3 cycle preclear with the -C switch on one of these drives here you are: EDIT: Using Bjp999's Unofficial Faster Preclear Script 133 Hours 19 Minutes and 52 Seconds ========================================================================1.15b == invoked as: ./preclear_bjp.sh -f -A -c 3 /dev/sde == ST8000AS0002-1NA17Z Z8404KRE == Disk /dev/sde has been successfully precleared == with a starting sector of 1 == Ran 3 cycles == == Using :Read block size = 1000448 Bytes == Last Cycle's Pre Read Time : 19:42:48 (112 MB/s) == Last Cycle's Zeroing time : 17:07:40 (129 MB/s) == Last Cycle's Post Read Time : 20:59:15 (105 MB/s) == Last Cycle's Total Time : 38:07:58 == == Total Elapsed Time 133:19:52 == == Disk Start Temperature: 24C == == Current Disk Temperature: 30C, == ============================================================================ ** Changed attributes in files: /tmp/smart_start_sde /tmp/smart_finish_sde ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Raw_Read_Error_Rate = 117 100 6 ok 126528792 Seek_Error_Rate = 75 100 30 ok 37790984 Spin_Retry_Count = 100 100 97 near_thresh 0 End-to-End_Error = 100 100 99 near_thresh 0 Airflow_Temperature_Cel = 70 76 45 near_thresh 30 Temperature_Celsius = 30 24 0 ok 30 Hardware_ECC_Recovered = 117 100 0 ok 126528792 No SMART attributes are FAILING_NOW 0 sectors were pending re-allocation before the start of the preclear. 0 sectors were pending re-allocation after pre-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 1 of 3. 0 sectors were pending re-allocation after post-read in cycle 1 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 2 of 3. 0 sectors were pending re-allocation after post-read in cycle 2 of 3. 0 sectors were pending re-allocation after zero of disk in cycle 3 of 3. 0 sectors are pending re-allocation at the end of the preclear, the number of sectors pending re-allocation did not change. 0 sectors had been re-allocated before the start of the preclear. 0 sectors are re-allocated at the end of the preclear, the number of sectors re-allocated did not change. ============================================================================
June 20, 201511 yr I have summarised the tests I performed in this thread in the Forum Sticky here: http://lime-technology.com/forum/index.php?topic=39526.msg386233#msg386233
June 20, 201511 yr Very nicely done Daniel -- nice to have the bulk of the data you've put together for us all summarized in one easy-to-read thread, without the need to ready through nearly 650 posts on 44 pages !!
June 30, 201511 yr Anyone tried 8TB HGST helium drives? I know its really expensive, but I would like to see the performance like come with Seagate 8TB.
June 30, 201511 yr Anyone tried 8TB HGST helium drives? I know its really expensive, but I would like to see the performance like come with Seagate 8TB. Ive been looking at them, but i think I'm just going to build a second server with 6TB HGST drives for now and wait for the price to come down some more.
June 30, 201511 yr Since these are still relatively new drives, I thought I'd post my preclear results for the 2 ST8000AS0002 I just installed: : ========================================================================1.15 : == invoked as: ./preclear_disk.sh /dev/sdb : == ST8000AS0002-1NA17Z Z8403xxx : == Disk /dev/sdb has been successfully precleared : == with a starting sector of 1 : == Ran 1 cycle : == : == Using :Read block size = 1000448 Bytes : == Last Cycle's Pre Read Time : 19:44:58 (112 MB/s) : == Last Cycle's Zeroing time : 16:22:29 (135 MB/s) : == Last Cycle's Post Read Time : 37:42:48 (58 MB/s) : == Last Cycle's Total Time : 73:51:20 : == : == Total Elapsed Time 73:51:20 : == : == Disk Start Temperature: 25C : == : == Current Disk Temperature: 37C, " : == : ============================================================================ : ** Changed attributes in files: /tmp/smart_start_sdb /tmp/smart_finish_sdb : ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE : Raw_Read_Error_Rate = 111 100 6 ok 35617992 : Seek_Error_Rate = 73 100 30 ok 21054924 : Spin_Retry_Count = 100 100 97 near_thresh 0 : End-to-End_Error = 100 100 99 near_thresh 0 : Airflow_Temperature_Cel = 63 75 45 near_thresh 37 : Temperature_Celsius = 37 25 0 ok 37 : Hardware_ECC_Recovered = 111 100 0 ok 35617992 : No SMART attributes are FAILING_NOW : : 0 sectors were pending re-allocation before the start of the preclear. : 0 sectors were pending re-allocation after pre-read in cycle 1 of 1. : 0 sectors were pending re-allocation after zero of disk in cycle 1 of 1. : 0 sectors are pending re-allocation at the end of the preclear," : the number of sectors pending re-allocation did not change. : 0 sectors had been re-allocated before the start of the preclear. : 0 sectors are re-allocated at the end of the preclear," : the number of sectors re-allocated did not change. : ============================================================================ : ========================================================================1.15 : == invoked as: ./preclear_disk.sh /dev/sdc : == ST8000AS0002-1NA17Z Z8403xxx : == Disk /dev/sdc has been successfully precleared : == with a starting sector of 1 : == Ran 1 cycle : == : == Using :Read block size = 1000448 Bytes : == Last Cycle's Pre Read Time : 19:52:10 (111 MB/s) : == Last Cycle's Zeroing time : 16:32:45 (134 MB/s) : == Last Cycle's Post Read Time : 37:49:29 (58 MB/s) : == Last Cycle's Total Time : 74:15:27 : == : == Total Elapsed Time 74:15:27 : == : == Disk Start Temperature: 26C : == : == Current Disk Temperature: 38C, " : == : ============================================================================ : ** Changed attributes in files: /tmp/smart_start_sdc /tmp/smart_finish_sdc : ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE : Raw_Read_Error_Rate = 112 100 6 ok 45860856 : Seek_Error_Rate = 73 100 30 ok 21308774 : Spin_Retry_Count = 100 100 97 near_thresh 0 : End-to-End_Error = 100 100 99 near_thresh 0 : Airflow_Temperature_Cel = 62 74 45 near_thresh 38 : Temperature_Celsius = 38 26 0 ok 38 : Hardware_ECC_Recovered = 112 100 0 ok 45860856 : No SMART attributes are FAILING_NOW : : 0 sectors were pending re-allocation before the start of the preclear. : 0 sectors were pending re-allocation after pre-read in cycle 1 of 1. : 0 sectors were pending re-allocation after zero of disk in cycle 1 of 1. : 0 sectors are pending re-allocation at the end of the preclear," : the number of sectors pending re-allocation did not change. : 0 sectors had been re-allocated before the start of the preclear. : 0 sectors are re-allocated at the end of the preclear," : the number of sectors re-allocated did not change. : ============================================================================
Archived
This topic is now archived and is closed to further replies.