July 8, 201213 yr I just ran a preclear on a brand new Seagate 7200 RPM 3 TB drive. At the end of the preclear, 3 values showed a "near threshold" status. - Spin retry count: New value = 100 | Old value = 100 | failure threshold = 97 | raw value = 0 - end to end error: New value = 100 | Old value = 100 | failure threshold = 99 | raw value = 0 - Airflow temperature: New value = 67 | Old value = 69 | failure threshold = 45 | raw value = 33 The drive temp hasn't gone above 33 as far as I saw during the whole preclear. I also noticed the pre-read and write speeds were much faster than the post-read. Is there any reason for this? I would have thought the pre and post read speeds would be about the same. Pre-read took 5:59 at 139 MB / Sec Zeroing took 6:49 at 122 MB / Sec Post-read took 12:53 at 69 MB / Sec Any thoughts? Anything to worry about? I was about to make this my parity drive. Pre-clear reports attached. preclear_finish__W1F0L5B0_2012-07-08.txt preclear_rpt__W1F0L5B0_2012-07-08.txt preclear_start__W1F0L5B0_2012-07-08.txt
July 8, 201213 yr Author The more I look at those first two, I'm guessing they just have a very low tolerance for any errors. I currently don't have any errors but the threshold is set such that it would fail even with only a few errors being recorded so it always shows "near threshold" Does this sound right? Any thoughts on the slowdown for post-read?
July 9, 201213 yr From my limited experience of pre-clearing a few drives over the weekend, all the drives I ran slowed down to a similar extent during the post-read. However, those values are worrying.
July 9, 201213 yr Author From my limited experience of pre-clearing a few drives over the weekend, all the drives I ran slowed down to a similar extent during the post-read. However, those values are worrying. I ran 3 more consecutive pre-clears. Will see what happens in a few days.
July 10, 201213 yr So I just finished running a 4-pass badblocks and 1 preclear on 2 3TB Seagate drives. Here are my errors on the 2 drives: old new raw read error rate = 117 119 seek error rate = 63 100 spin retry count = 100 100 end to end error = 100 100 airflow temperature cel 61 63 temperature celsius 39 37 no smart attributes are failing now no sectors were re-allocated or are pending re-allocation raw read error rate 117 119 seek error rate 63 100 spin retry count 100 100 end to end error 100 100 airflow temperature cel 62 64 temperature celsius 38 36 no smart attributes are failing now no sectors were re-allocated or are pending re-allocation So from this, I think certain attributes are inherent in these 3 TB drives where a laser heats up the disk in order to increase the density. I wonder if the heat is related. There seems to be some mechanical difficulty related to this, but in the end, it passed 4 badblocks runs and a preclear without any errors. I have 4 other identical drives currently going through badblocks. I'll see what happens after the preclears, but I feel they will end up looking the same.
July 14, 201213 yr Author Well here are my final results after 3 additional preclear passes. Disk /dev/sdd has been successfully precleared == with a starting sector of 1 == Ran 3 cycles == == Using :Read block size = 8225280 Bytes == Last Cycle's Pre Read Time : 5:59:38 (139 MB/s) == Last Cycle's Zeroing time : 6:48:50 (122 MB/s) == Last Cycle's Post Read Time : 12:52:04 (64 MB/s) == Last Cycle's Total Time : 19:41:56 == == Total Elapsed Time 65:23:25 == == Disk Start Temperature: 30C == == Current Disk Temperature: 29C, == ============================================== ** Changed attributes in files: /tmp/smart_start_sdd /tmp/smart_finish_sdd ATTRIBUTE NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS RAW_VALUE Raw_Read_Error_Rate = 115 116 6 ok 95850784 Seek_Error_Rate = 66 61 30 ok 5097988 Spin_Retry_Count = 100 100 97 near_thresh 0 End-to-End_Error = 100 100 99 near_thresh 0 High_Fly_Writes = 97 100 0 ok 3 Airflow_Temperature_Cel = 71 70 45 ok 29 Temperature_Celsius = 29 30 0 ok 29 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. I'm feeling pretty confident the drive is fine. I did some reading online that discussed these values in other model Seagate drives and I think it has to do with how Seagate chooses to calculate and track those numbers. Here is one of the links if interested. http://forums.seagate.com/stx/board/crawl_message?board.id=ata_drives&message.id=25651 And also found this excerpt from another site about the end to end error: fzabkar wrote: AFAIK, HP is responsible for introducing the End-to-End Error attribute, but it is not responsible for introducing SMART. That was Compaq's idea: http://en.wikipedia.org/wiki/S.M.A.R.T. A raw value of 0 for the End-to-End Error attribute indicates that the drive has recorded no such errors. This means that this particular attribute is perfectly healthy, in which case it receives the maximum possible "health score". This is reflected in the normalised value of the attribute, ie 100. Each error reduces the normalised value by 1. Therefore a threshold value of 99 would suggest that 1 error is enough to fail this attribute. In short, an "adjusted" value of 100 is a perfect score that corresponds to zero errors. If you examine other SMART reports in this thread, you will see an example where the normalised value is 96 and the raw value is 4. Once again this confirms that each error results in the loss of one "health" point. Third party SMART tools all report exactly the same data as SeaTools. However, where they differ is that they also report the raw values whereas SeaTools does not. Where these third party tools go wrong is when their authors misinterpret the raw values. For example, the Seek Error Rate and Raw Read Error Rate attributes create a great deal of unwarranted concern because their numbers are counterintuitively large.
July 15, 201213 yr From my limited experience of pre-clearing a few drives over the weekend, all the drives I ran slowed down to a similar extent during the post-read. However, those values are worrying. You are right, you have limited experience. (but I see you are rapidly gaining in knowledge) Nothing is unusual. The post-read is actually "post-read AND verify". It will take about twice as long as the pre-read phase. Yes, on some parameters, the manufacturer sets the failure threshold just a few counts from the normalized starting value. Usually it meams a few failures of that type indicates a problem that will fail smart. You cannot know however... the manufacturers do not tell us how many actual failures resent one increment of a normalized value. Do not confuse normalized value with raw values. Only on some RAW values are the counts actual counts, and even then there is very little standardization. usually, the RAW temperature is in degrees, but sometimes it has an offset. Usually un-readable sector counts are actual counts, as are re-allocated sectors.
July 15, 201213 yr In short, an "adjusted" value of 100 is a perfect score that corresponds to zero errors. Sometimes... Sometimes it is 200, sometimes 253. It completely depends on the manufacturer/drive model. Sometimes a normalized value will start as 253, but as soon as the drive is put into service changes to 100, or 200, as its beginning health value. The irritating thing about the "SMART" standard is that the manufacturers each use their own "standard" for any given disk.
July 16, 201213 yr Author In short, an "adjusted" value of 100 is a perfect score that corresponds to zero errors. Sometimes... Sometimes it is 200, sometimes 253. It completely depends on the manufacturer/drive model. Sometimes a normalized value will start as 253, but as soon as the drive is put into service changes to 100, or 200, as its beginning health value. The irritating thing about the "SMART" standard is that the manufacturers each use their own "standard" for any given disk. I have noticed and it really can be a pain. Whenever I see a smart value that doesn't make sense I start with a web search for that manufacturer and smart value. Seems they all run there number differently and here is usually nothing to worry about.
Archived
This topic is now archived and is closed to further replies.