This message is from the T13 list server.
Any SATA experts care to explain what happened?
BACKGROUND
* A name brand motherboard chipset that supports SATA - used for nearly a year of testing PATA and SATA devices with no strange problems - certainly no problem like the problem described here.
* Two name brand SATA hard disk drives that have never failed to execute a valid ATA command correctly. Both system and drives have been used for something over 1000 hours of write/read testing with no problems like the problem described here.
* The system and the drives had not been powered on for several weeks.
CONFIGURATION
Drive "A" attached to SATA port 1. Drive "B" attached to SATA port 2. Both SATA ports on the motherboard.
TEST RESULT (THE FAILURE)
1) The system (with drives attached) was powered on.
2) A basic ATA write/read test was started on Drive "A" (SATA port 1).
3) The test immediately fails on WRITE SECTORS commands with Sector Counts or 3, 5 and 7. In each case, after the SATA controller and drive
transferred the proper number of sectors (3, 5 or 7) the controller would claim the drive had status of BSY=0 DRQ=1 indicating the drive wanted the host to transfer another sector. WRITE SECTORS commands with Sector Counts of 2, 4 and 6 work OK.
4) Cycle system power off and on.
5) Repeat 2) with same result as described by 3).
6) Start same basic ATA write/read test on Drive "B" (SATA port 2).
7) Drive "B" (SATA port 2) runs test with no errors.
8) Power system off, swap the SATA cables at the drive connectors so that Drive "A" is now on SATA port 2 and drive "B" is now on SATA port 1.
9) Power on system and repeat the test on Drive "A" (SATA port 2) - no error. Repeat test on Drive "B" (SATA port 1) - fails as described in 3). THE FAILURE STAYS WITH THE SATA PORT/CABLE!
10) Turn off system, swap SATA cables back so that Drive "A" is on SATA port 1, Drive "B" is on SATA port 2 (back to original configuration).
11) Power on system - test Drive "A" - no error - test Drive "B" - no error.
WHAT HAPPENED???
Please!!! Someone explain this!
Is this another of those undetected "data errors" in a command FIS - an error that corrupts the data in one or more of the ATA Command Block registers as the FIS is sent to the drive? But why just on SATA port 1? And why only for a short period of time (something like 5 minutes)?
It is strange and unreproducible errors like this that make people say SATA is unreliable. This is bad... this is really bad.
BTW the system and the two drives have been running with no errors for 3 days since this problem happened and then magically "disappeared".
Hale
--
++ Hale Landis ++ www.ata-atapi.com ++
