|
Hi General,
I am not too smart about hardware but...
I was looking this up all last week.
I was getting these messages:
<4>hdb: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
<4>hdb: read_intr: error=0x10 { SectorIdNotFound }, LBAsect=19606367, sector=19541088
<4>end_request: I/O error, dev hdb, sector 19541088
<3>Buffer I/O error on device hdb, logical block 19541088<4>hdb: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
<4>hdb: read_intr: error=0x10 { SectorIdNotFound }, LBAsect=19606367, sector=19541088
<4>end_request: I/O error, dev hdb, sector 19541088
<3>Buffer I/O error on device hdb, logical block 19541088
I read, amongst my searches, such advice as "your drive is stuffed and about to die" but I didn't believe it because I have been getting these messages about hdb since I installed Linux over a year ago. Also, I have got similar messages about hda since then, and it was brand new at the time. Both disks are still working fine.
It is still all a puzzle to me, but experts might know what the codes mean where it says "error=0x10" and "status=0x59".
(If you are getting other numbers in these messages it might be more worrying.)
The "sector=1954108" mentioned happens to be the last sector on that drive.
On boot, there would be these messages, then DMA would be turned off.
If I later turned DMA back on for hda, the system would find this problem with hdb again at some time and turn off DMA for hda and hdb (which are on the same DMA channel.
However if I "force DMA", i.e. turn on DMA for all the disks at a later stage of booting, the problem does not happen any more: I don't get the messages, DMA is not turned off.
If you want to search for information, I found that searching for "SeekComplete DataRequest Error" turned up plenty of results: but it is still a mystery to me. On some forums I saw discussions that seemed to me to be among hardware experts -- I didn't really understand a word of it.
I concluded that, at least with these particular error codes -- "error=0x10" and "status=0x59" -- the cause is probably in the areas of DMA and ACPI -- and the worst that will happen is that DMA will be turned off while you are working (annoying!).
(By the way, in case it is relevant, I found out in these discussions that if you have power saving enabled on your linux system, a disk that goes to sleep is likely to have DMA still disabled when it is woken up.)
Like you said, I saw that other people had run disk checking utilities after seeing these messages -- and those checks didn't show up any problems with the disk.
Also I noticed that at shutdown, i.e. runlevel 6, a similar message is returned about every disk. This makes sense to me, i.e. they are unmounted now or turned off, so they are not returning any info to the system.
Hope this helps but, as I said, after a lot of searching I found that quite a few people have questions about these messages but I still did not find out anything aimed at my level of understanding.
|
