Messages in this thread | | | Date | Tue, 12 Mar 1996 14:12:34 -0600 (CST) | From | Simon Shapiro <> | Subject | Re: SCSI Kernel Problem - BAD |
| |
Hi Alan Cox; On 12-Mar-96 you wrote: > > > problem is more severe when the I/O rate is high. not the amount of data > > as much as the number of I/O operations per second. In a PC, these are > > typically directly related to interrupts. I also noticed that a busy NFS > > connection will aggravate it even more. and I just saw that when NFS was > > busy on these partitions, I was loosing interrupts on the wd.c driver. > > Heavy NFS load significantly increases the probability of GFP_ATOMIC or > GFP_DMA memory allocations failing. There are also other things that will > start to happen at this load (clustering attempts fail, the request queue is > full, buffer cache grow/shrink). Is it worth trying these one at a time - > eg a 4 item request queue and seeing if you can break it. > > Alan > Thanx ( and Hi! ) Alan. I can reproduce the problem easily without any nfs activity. As I said elsewhere, it does not smell of queue (or other resource) overflow. I can load the machine to load average of 4-6 and it will be fine. Do something it hates, on an idle machine - Poof!
Sincerely Yours, (Sent on 03/12/96, 14:12:34) Simon Shapiro i-Connect.Net, a Division of iConnect Corp. Shimon@i-Connect.Net 13455 SW Allen Blvd., Suite 140 Beaverton OR 97008
| |