2.4.26: IDE drives become unavailable randomly
I was getting this problem, and advice from smartmontools people was to clean out the box and reseat all cables etc. Seemed to work for me on the box at work with this DMA timeout issue - BTW, always happened at idle, like 2:15am in the middle of the night etc.
Reference:
http://sourceforge.net/mailarchive/message.php?msg_id=8660397
http://sourceforge.net/mailarchive/forum.php?thread_id=4908273&forum_i
d=12495
Nick
Signature · 2 lines
"When you're chewing on life's gristle, Don't grumble, Give a whistle..."
Quoted 9 lines
I was getting this problem, and advice from smartmontools people was to clean out the box and reseat all cables etc. Seemed to work for me on the box at work with this DMA timeout issue - BTW, always happened at idle, like 2:15am in the middle of the night etc.
Reference:
http://sourceforge.net/mailarchive/message.php?msg_id=8660397
http://sourceforge.net/mailarchive/forum.php?thread_id=4908273&forum_i
d=12495
An additional reference. See the entry that starts 'System freezes under heavy load" in:
http://cvs.sourceforge.net/viewcvs.py/smartmontools/sm5/WARNINGS?view=markup
Cheers,
Bruce
(please cc me on any replies, I am not subscribed to this list)
Quoted 10 lines · On Thu, 1 Jul 2004 04:48:30 -0500 (CDT) Bruce Allen…
Quoted 9 lines
I was getting this problem, and advice from smartmontools people was to clean out the box and reseat all cables etc. Seemed to work for me on the box at work with this DMA timeout issue - BTW, always happened at idle, like 2:15am in the middle of the night etc.
Reference:
http://sourceforge.net/mailarchive/message.php?msg_id=8660397
http://sourceforge.net/mailarchive/forum.php?thread_id=4908273&forum_i
d=12495
Quoted 6 lines
An additional reference. See the entry that starts 'System freezes under heavy load" in:
http://cvs.sourceforge.net/viewcvs.py/smartmontools/sm5/WARNINGS?view=markupCheers,
Bruce
Thks, folks, I wouldn't really suspect of bad cables/PSU, this was an eye-opener. I have just opened the box and reseated the 80-wire IDE cable to my hda device, and I will consider replacing it, just in case. The PSU is brand new, 450W -- although it could be bad quality, I will try to check this out.
BTW: Nick, I missed your msg because you didn't cc me. My hda also usually gets disconnected at early hours in the morning, as you pointed out. I arrived today to work and it had happened again =/ Last entry on/var/log/messages was around 1:30am, and it was about a NFS mount that had expired.
Best,
Andre
Signature · 2 lines
Andre Oliveira da Costa ([email protected])
Quoted 34 lines
(please cc me on any replies, I am not subscribed to this list)
Quoted 10 lines · On Thu, 1 Jul 2004 04:48:30 -0500 (CDT) Bruce Allen…
Quoted 9 lines
I was getting this problem, and advice from smartmontools people was to clean out the box and reseat all cables etc. Seemed to work for me on the box at work with this DMA timeout issue - BTW, always happened at idle, like 2:15am in the middle of the night etc.
Reference:
http://sourceforge.net/mailarchive/message.php?msg_id=8660397
http://sourceforge.net/mailarchive/forum.php?thread_id=4908273&forum_i
d=12495Quoted 6 lines
An additional reference. See the entry that starts 'System freezes under heavy load" in:
http://cvs.sourceforge.net/viewcvs.py/smartmontools/sm5/WARNINGS?view=markupCheers,
BruceThks, folks, I wouldn't really suspect of bad cables/PSU, this was an eye-opener. I have just opened the box and reseated the 80-wire IDE cable to my hda device, and I will consider replacing it, just in case. The PSU is brand new, 450W -- although it could be bad quality, I will try to check this out.
BTW: Nick, I missed your msg because you didn't cc me. My hda also usually gets disconnected at early hours in the morning, as you pointed out. I arrived today to work and it had happened again =/ Last entry on/var/log/messages was around 1:30am, and it was about a NFS mount that had expired.
Quoted 3 lines
Best,
Andre
Hi Andre,
Sorry, I too am not subscribed to the list, and I read (and reply to) from:
http://marc.theaimsgroup.com/?l=linux-kernel
I totally overlooked CC'ing you.
Anyway, new IDE cable did fix the box at work for me. Also I only used smartd AFTER the problem arose, not before, so it was not smartd that caused it.
Regards,
Nick
Signature · 2 lines
"When you're chewing on life's gristle, Don't grumble, Give a whistle..."
(please cc me on any replies, I am not subscribed to this list)
Hi Nick,
On Thu, 01 Jul 2004 19:43:58 +0100
"Nick Warne" <[email protected]> wrote:
[snip]
Quoted 16 lines
Quoted 12 lines
Thks, folks, I wouldn't really suspect of bad cables/PSU, this was an eye-opener. I have just opened the box and reseated the 80-wire IDE cable to my hda device, and I will consider replacing it, just in case. The PSU is brand new, 450W -- although it could be bad quality, I will try to check this out.
BTW: Nick, I missed your msg because you didn't cc me. My hda also usually gets disconnected at early hours in the morning, as you pointed out. I arrived today to work and it had happened again =/ Last entry on/var/log/messages was around 1:30am, and it was about a NFS mount that had expired.
Quoted 3 lines
Best,
Andre
Quoted 8 lines
Hi Andre,
Sorry, I too am not subscribed to the list, and I read (and reply to) from:
http://marc.theaimsgroup.com/?l=linux-kernel
I totally overlooked CC'ing you.
No worries. I have to remind myself over and over again to put the "please cc me" header on every msg =)
Quoted 3 lines
Anyway, new IDE cable did fix the box at work for me. Also I only used smartd AFTER the problem arose, not before, so it was not smartd that caused it.
Thks, this is definitely valuable info. I will surely go after some new cables to see if it improves my situation (I hope it does, otherwise I will have to move services away from this box and relocate them to other servers, which will be a PITA...)
On a side note, I browsed 2.4.27rc2 changelogs today, and there is some interesting stuff there about interruptions, ACPI etc. Looking forward to trying it out.
Thks again,
Andre
Signature · 2 lines
Andre Oliveira da Costa ([email protected])