Joe L.

You might be right... and I could only test the -t on a simulated disk here, so it too could be an issue.. but not as likely.

I did post it a page back in this thread... Sorry you did not notice.

I cannot tell either... as I have no 3TB drive to test with. Too bad you did not run this command before assigning the disk to the array: After invoking it, and before assigning the drive to the unRAID array, you can post the results of this command to assist me in knowing if it worked as expected on the 3 TB drive: dd if=/dev/sda count=1 2>/dev/null | od -x -A x (substituting your disk for /dev/sda ) It should have returned 0000000 0000 0000 0000 0000 0000 0000 0000 0000 * 00001c0 0002 ff00 ffff 0001 0000 ffff ffff 0000 00001d0 0000 0000 0000 0000 0000 0000 0000 0000 * 00001f0 0000 0000 0000 0000 0000 0000 0000 aa55

Yes, the -t should work on large disks. (I'm guessing it is no longer a valid pre-clear signature if it "converted" it)

Running at 143 MB/sec now @ 7% complete. Not sure I'll be home when it gets to the inner cyl, but will have to see. BTW, this is a 7200 RPM Hitachi drive. If it can keep up that speed it will be 6 or 7 hours for the "Write" phase. Figures though, since the rotational speed of the disk is the same, and the density on the platters higher, the write speed should also be higher. I expect you'll be down to 100MB/s or less on inner cylinders... still very nice.

Yes, I have just finished one... I'll attach it to this post temporarily. (Until it gets tested by a few brave users) It has NEVER been tested on an actual 3TB drive... It should work. The GPT signature it creates is identical to the one Tom at lime-technology described to me. You'll have to give it a try and let me know how it goes. It can not convert GPT partitions using the "-C" option, as they are handled entirely differently and never start on partition 63 or 64. Also, if your disk is over 2.2TB in size, it will set the preclear signature as it needs, completely ignoring any "-a" or "-A" option you might try to give it. Joe L. Running now, with the "-W" option (disk was precleared previously and has been running in 2.2T mode for a while). Zeroing at 140 MB/sec The initial screen is still saying stuff about partition alignment / offsets, etc. It should probably say it is a >2TB disk and being prepared for a GPT structure. My disk already had an MBR from the 2.2T mode (all I did was remove the HPA, reboot, and start this preclear). So this should be a pretty good test of preclear's ability to preclear a 3T drive. Oops... forgot to change the initial screen and what it says when a GPT preclear will occur. It is actually a bit tricky, since the MBR partition would STILL be used by unRAID if the disk advertised its size in 4K blocks as opposed to advertising its size in 512 byte blocks. Only if the size is >= 0xFFFFFFFF blocks(sectors) is a GPT partition used by unRAID. The math work on the number of blocks, not the disk size in bytes. I'll be curious to see how it does. What kind of "write" speed is it showing? (If I would only read I'd know. Will be curious to see speed on inner cylinders) Joe L.

Yes, there is a option to run a short-test. However, it purposely does NOT write a valid pre-clear signature. (again, to prevent somebody from attempting to pre-clear using it thinking it is doing what is needed to protect their parity data) I'll send you a PM describing how to invoke it if you like... After invoking it, and before assigning the drive to the unRAID array, you can post the results of this command to assist me in knowing if it worked as expected on the 3 TB drive: dd if=/dev/sda count=1 2>/dev/null | od -x -A x (substituting your disk for /dev/sda )

nope, no mode to fake a pre-clear signature. That is by design. Too likely to be mis-used and destroy somebody's parity integrity. Formatted does not mean pre-cleared, in fact, it is EXACTLY the opposite. A formatted drive has tons of data on it already from the formatting process. Only shortcut is to use the "-n" option to skip the pre and post read phases. That will zero the drive and write the pre-clear signature, but it WILL NOT identify bad sectors that can cripple parity re-construction of a failed drive later, when you learn they are un-readable. Use of the "-n" option will result in the drive being zeroed at about a 100MB/s rate on most modern hardware. 10 seconds for 1Gig, 3000Gig = 30000 seconds = roughly 8.33 hours. When done, I have no idea if unRAID will tell you anything when you add the drive to an existing array to let you know it is pre-cleared (It would be nice if it did though). It will either add the drive quickly, or it will proceed to clear the drive on its own, and take your array off-line for the same 8.33 hours while it does. The "-t" option will tell you if the pre-clear script thinks the drive is pre-cleared. I'd use it first, and if it says the drive is pre-cleared, you are probably ok. Joe L.

Yes, I have just finished one... I'll attach it to this post temporarily. (Until it gets tested by a few brave users) It has NEVER been tested on an actual 3TB drive... It should work. The GPT signature it creates is identical to the one Tom at lime-technology described to me. You'll have to give it a try and let me know how it goes. It can not convert GPT partitions using the "-C" option, as they are handled entirely differently and never start on partition 63 or 64. Also, if your disk is over 2.2TB in size, it will set the preclear signature as it needs, completely ignoring any "-a" or "-A" option you might try to give it. Joe L. preclear_disk1.12beta.sh.zip

They looked fine. You can try another pre-clear. Basically, after being cleared, the whole disk should be read back as zeros. If not, something is wrong. Of course, it could be bad memory or a bad power supply, or a bad disk controller... no way to know until you isolate it. These types of errors cause hair loss, and random parity errors as disk returns different than expected. If it is the disk, it is best next tested as a wheel chock. (You can test how effective it is in impeding the forward and backward motion of your car as you drive over it repeatedly ) You are welcome.

/dev/sdh did not do too good. 23 sectors are pending re-allocation at the end of the preclear, a change of 23 in the number of sectors pending re-allocation. Run another pre-clear on it, and if the number of sectors pending re-allocation does not go to zero, consider an RMA for it too.

Looks like you might be using an older version of the script?, the -A option was only added in recent updates. exactly. I'll bet it is 10 or more versions old. what do you get when you type preclear_disk.sh -v

Looks like you might be using an older version of the script?, the -A option was only added in recent updates. exactly. what do you get when you type preclear_disk.sh -v

Hello. I have Windows and I'm editing the Go script via Putty/Telnet using Midnight Commander. Is there a command or mouse click that I can do to copy the beep codes above and paste it in Midnight Commander? In "putty" a right hand mouse click pastes what is in the clipboard. In windows, "control-c" copies what is highlighted to the clipboard.

You can as long as you are not using 5.0beta6a. There is a bug in that version that will refuse to identify a multi-partitioned cache drive as valid, so it re-partitions it in error back to a single partition. For earlier versions, unRAID will use the 1st partition for cache, leaving the others for your use. Joe L.

unless you have an absolute need for a 2.5 inch drive, I'd get one of the 2TB drives for only a small amount more money. $59, free shipping, 1.5TB http://www.newegg.com/Product/Product.aspx?Item=N82E16822148725&cm_sp=DailyDeal-_-22-148-725-_-Homepage $59 after rebate, $69 if you forget to send in the rebate form, free shipping 2TB Hitachi... (one of my favorite brands) http://www.newegg.com/Product/Product.aspx?Item=N82E16822145475 http://www.newegg.com/Product/Product.aspx?Item=N82E16822145475

That's too hot, especially for idle. Is it a 7200rpm drive? A lot of the older drives ran very hot. It might be normal for that drive to consume a lot of power, and waste a lot of it in heat. Probably just needs a lot of air-flow across it. (A 36" fan, spaced about an inch or two away should do it... If it does not, just use multiple ) I'll bet the older drive is not "green"

Based on what you said, I'd RMA the drive. Last thing you want is one that constantly has problems reading sectors.

The power cycles are over then entire life of the drive, not just the pre-clear time. Here is an example from one new 2TB Hitachi drives recently pre-cleared 9 Power_On_Hours 0x0012 100 100 000 Old_age Always - 224 10 Spin_Retry_Count 0x0013 100 100 060 Pre-fail Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 5 192 Power-Off_Retract_Count 0x0032 100 100 000 Old_age Always - 5 193 Load_Cycle_Count 0x0012 100 100 000 Old_age Always - 5 and here is one from a 500Gig Hitachi drive that has been in operation for much longer in my older server (At the time, it was the largest you could buy). 9 Power_On_Hours 0x0012 094 094 000 Old_age Always - 47780 10 Spin_Retry_Count 0x0013 100 100 060 Pre-fail Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 274 192 Power-Off_Retract_Count 0x0032 097 097 050 Old_age Always - 4757 193 Load_Cycle_Count 0x0012 097 097 050 Old_age Always - 4757 47780 hours = nearly 5 1/2 years. Looks like the oldest drives in my array are now out of warranty. I hope our new drives last as long... Joe L.

First step is to ensure your power supply is up to the task of powering all your disks. Please list the exact make model and the number of disks connected to it. Every "read" phase seems to identify un-readable sectors. The subsequent "write" phase seems to re-write them in place, but then the post-read cannot read them again, and again they are marked as pending re-allocation. If it were me, and the disk was connected to a good power supply, I'd RMA it.

If it formatted the drive, it is no longer cleared. You can test with preclear_disk.sh -t /dev/sdX It will tell you if it thinks it is cleared.

You can ignore almost all of that in the smart report. There is one re-allocated sector, none pending re-allocation. Unless you know something we do not, it is fine. Run it through a preclear cycle. If no more sectors show up as being re-allocated, then use the drive. Joe L.

preclear_disk.sh -v

Probably can detect the failure and try without the "-d ata" Not a bad idea for the next version. Thanks for the feedback. How are you Joe L., two questions, did the above make it into any of the releases after 1.9 (1.10 or 1.11)? If i am currently running a preclear (using version 1.9) can I update the preclear_disk.sh file on my key to the lastest version, or is it in use and not a good idea. Or just not a good idea at all while a preclear cycle is currently running? I did not yet add the logic, so your version is as good as any. The big difference between 1.9 and 1.11 is the ability to monitor the progress on the MyMain screen. 1.10 was never released, it was an internal version number. Do not change the script while it is running. Just let the current pre-clears you have running complete. Thanks for the reminder though, I had not been thinking of the change to add that logic much... Joe L. Thanks Joe L. I will hang tight until it finishes before upgrading to v1.11, another question (I may have missed it if it was already stated) If you are preclearing more than one drive will the results for all (each) of the drives display in the MyMain page OR at this time the design is only capable of displaying the results of one drive preclearing? It will show multiple. I would change it to device_type=""

Probably can detect the failure and try without the "-d ata" Not a bad idea for the next version. Thanks for the feedback. How are you Joe L., two questions, did the above make it into any of the releases after 1.9 (1.10 or 1.11)? If i am currently running a preclear (using version 1.9) can I update the preclear_disk.sh file on my key to the lastest version, or is it in use and not a good idea. Or just not a good idea at all while a preclear cycle is currently running? I did not yet add the logic, so your version is as good as any. The big difference between 1.9 and 1.11 is the ability to monitor the progress on the MyMain screen. 1.10 was never released, it was an internal version number. Do not change the script while it is running. Just let the current pre-clears you have running complete. Thanks for the reminder though, I had not been thinking of the change to add that logic much... Joe L.