On Thu, 2008-03-20 at 16:17 -0500, James Bottomley wrote: > On Thu, 2008-03-20 at 22:08 +0100, Raoul Bhatia [IPAX] wrote: > > hi james, > > > > James Bottomley wrote: > > > A work around might be to lower the queue depth to say 4 or 8 and up the > > > retries (this latter can only be done by altering the SD_MAX_RETRIES > > > parameter in include/scsi/sd.h and recompiling). > > > > any suggestions for these parameters: > > > > include/scsi/sd.h: > > > #define SD_TIMEOUT (30 * HZ) > > > #define SD_MOD_TIMEOUT (75 * HZ) > > > > > > #define SD_MAX_RETRIES 5 > > > #define SD_PASSTHROUGH_RETRIES 1 > > > > shall i alter the timeouts? > > The timeouts can be altered on the fly > at /sys/class/scsi_device/<device>/device/timeout > > but I'd leave them as is for now ... it wasn't the timeouts that fired. > > > what is a good value for the retries? time it by 2? by 5? by 10? > > I'd up the retries to say 15 and see if that works. > > > moreover, where can i lower the queue? > > echo <new depth> > /sys/class/scsi_device/<device>/device/queue_depth We have played a lot with the queue depth on this controller. So far the best we have done is extended the time before an end device is eventually dropped. I am very curious to see what happens in Raoul's test case. > > James > > > -- > To unsubscribe from this list: send the line "unsubscribe linux-scsi" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
