Messages in this thread |  | | | Date | Wed, 20 May 1998 10:21:15 -0400 | | From | Dave Cinege <> | | Subject | Re: 2.0.34pre15 opps.....DAMN IT! |
| |
Alan Cox wrote: > > > For starters it a the DPT-Quantum fireball combo and I would > > Current quantum fireball firmware ? Several fireballs had tagged queueing > bugs where the drive got bored of talking SCSI if I remember rightly
Vendor: QUANTUM Model: FIREBALL ST4.3S Rev: 0F0J
I got this firmware update from Quantum tech support about a month ago. It did not help.
Peter Svensson wrote: > > The most current firmware revision also have a problem with a tagged-queue > depth of more than 7 under heavy load.
Though it will die from load, it normally goes into EATA resets during pretty light load as if it has a 'wakeup' problem.
I'm planning to replacing the drive soon. The thing worring me is the opps. I'm seeing the opps more then the resets right now.
> > occasionaly see EATA-Resets. I was replaced the mobo and that > > seemed to produce Opps and a higher degree of EATA-RESETS. The 3 > > other boxes I have with this same board run like a rock (but run > > OS/2 and DRDOS, not linux) > > If the changed board is the only one that goes Oops Id guess you swapped > a good for bad board ?
Just before the board change happened the system was put into full service. I *might* have seen an opps with the old board. Before that the system was running for several months with only the occational eata reset...but of course this was with practically no load.
FYI the old board was an ABIT IT5H. Intel HX chipset. This one had a Cyrix P200MX.
Also FYI, a system sitting right next to it is a dual P100, with a 3.2GB quantum fireball. Tyan Titan 3 96MB. Hell of a time getting memory above 64MB working but once I did it runs very good. I DO see EATA resets on it, but they are so rare it's not a major concern. Right now this box has been up 9 days and has assumed the load that should be going to the other box.
I have swapped the DPT's around. No change.
> Is that board one of those that overclocks the PCI bus to 37.5MHz when you > run 75MHz bus ?
Yes it runs the PCI sync/2. (Which is the point to the faster bus IMO) The PCI bus contains 2 SVEC PN100TX and right now a DPT 2144W. This DPT was previously in my personal system up until a few months ago. That has a custom 90MHz bus (45MHz PCI) and the HBA ran flawlessly for over a year. The Aladdin chipset is rated to 83MHz. I doubt any of this is brought on by bus speed.
One thing I left out of the opps report....I'll see stuff like this: free_one_pmd: bad directory entry 01000000 in dmesg sometimes. Once that's present the box has a few hours left in it at best before an opps.
-- http://www.psychosis.com/emc/ Elite MicroComputers 732-541-4214 http://www.psychosis.com/linux-router/ Linux Router Project
"The only way to remain cutting edge, is to design for things that do not yet exist." -- Me : )
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
|  |