Seagate Barracuda 7200.11 HD Owners - Must read!

Necrosis

Neshi you also might want to "scroll down" in Everest.

Neshi

ah darn... I thought I scrolled down enough. apparently not.
that's a stupid mistake to make. It is displaying 32mb buffer..
*slaps himself on the back of the head*

Necrosis

Ha, ha don't worry I did the same thing, have the 750GB.

RealHJ2

I have a question regarding the "Sandra" reporting and overall speed on the Segate 1TB ST31000340AS SAS drives.


Brief background: I have a Gigabyte P35-DS3R based pretty good system (8GB Crucial top-grade RAM, Q6600 CPU, Intel ICH8R/ICH9R SATA RAID controller on-board, etc.). The only problem with it: Windows Vista (I should have never listed to the person at work who convinced me to get it w/ chicken Windows vs. Win XP).


I recently put in a pair of 1TB ST31000340AS'es, firmware SD15. So, presumably, it should not be affected by this specific problem.


However, two issues that would lead me to believe that perhaps it is affected:


1.a) Super slow write speeds to the Seagate 1TB drive. At most, I can get some 4 MB/s consistent writes to it (this is to a single drive, putting data on it before building RAID-1 off it). This is many, many times slower than I get now on the WDC WD-5000ABYS-01TNA0 500GB drives I have in another RAID-1 (the Seagate 1TBs are to replace the second 500GB RAID-1 with a 1TB RAID-1). To Copy some 450GB of data I first needed to do it in small chunks (anything over 16,000 files copied at once and chicken windows (Windows Vista) will just choke and some 15-20 minutes into the file copy just freeze, all goes white, system unresponsive - as it can't handle more than 16,000 (approx.) file copy at once - if any more attempted, need to hard reboot once chicken windows chokes), and overall it took about a day and a half to copy the 450GB over. I measured approx 6 hours to copy each 100GB of data.


b) Building RAID-1 from one 1TB drive to the other it has taken some 20 hours and so far it's only 58% done in the "mirroring" process. So, it would seem that it takes some 36 hours - 1 and a half days! - just to mirror some 450GB of data from one Seagate 1TB drive to the other. This is not copying file by file, but a low level mirror by the Intel ICH8R/ICH9R RAID controller. The same task on my 500GB WD drives RAID-1 would only take approx an hour and a half, I know.


2. Sandra reports the Seagate 1TB drive cache differently every time I look at it. It ranges from 16KB to 10.xxMB. (Though in general it seems not accurate on that reporting anyhow for all the drives. Is there any better, more reliable and consistent reporting tool recommended?)


3. Another item: the Seagate 1TB drive seems to chronically overheat. More than once during the 450GB data copy the drive just "disappeared" and the system would not see it. Then the only solution is to shut system down and some 10-15 min later reboot and it'll see (rebooting right away it won't be seen by the bios). It seems that these Seagate 1TB drives are very low-end consumer grade and are not meant for continous I/O operations - as anything over an hour consistent reads or writes and the drive is at a risk of overheading and (I presume) auto shutting down due to overheat. (Note: at home I have AC set to 22C consistently, it is not hot here. And, the Seagate 1TB drives in a chassis are set with the 5.25" adapter brackets, so there is lots of airflow above and below - they are not squished tight together in the 3.5" drive space, but instead have ample cooling and good fans, etc. in the system.)


Any recommendations re: the SD15 firmware? Is it also known to have the cache problem? Should I update to AD14? Would that fix the super-slow reads and writes of these drives? (Note: I set jumper to "Generation 2" - SATA 300 - so it's not that it's set to SATA 150 slower default.) And/or any other suggestions regarding the slow reads/writes by these drives and the chronic overheating when using the drive for 1+ hour continous reads/writes (as is needed to move the 450GB data to it and then build the RAID-1 mirror off the drives)?

Judas

That sounds like something critically wrong with the drive itself.... i can't see vista having anything to do with the problem at all...

have you run Seagate Tools to check the drive itself....

even the horrible PIO MODE 2 hd's i had on hand produced a little better results then that.

check your sata cable as well.. or the connection... something is surely wrong.

RealHJ2

It's not the SATA connection for sure, as it's using the same ports (4,5) and the same SATA cables as my earlier 500GB WD drives had, which had about 20 times faster reads & writes as these drives do. It's highly unlikely that just by unplugging the SATA cables from the earlier two drives and plugging into these the two drives, both SATA cables would have went bad.

I just run the Seagate SeaTools. Short drive self test always fails, SMART test unavailable, etc... so no real conclusive results.
But, again, it doesn't see the drives: it only sees the RAID arrays, so that is perhaps why as the Intel RAID controller is blocking SeaTools from seeing the drives themselves directly.. I'll try the DOS version of the tools intead perhaps now.

Judas

If the short test EVER fails.. the drive is bad..

And yes you'd like having to "unraid" the drives to do it... but i don't think it should fail any tests...