SUMMARY-->Panic : Error In Swapping in U-Area

From: Scott A. Surguine (surguine@densun0.asa.org)
Date: Thu Feb 11 1993 - 14:46:16 CST


Greetings Fellow Sun Managers:

Several Posts regarding potential Disk errors over the past week:

>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: SCSI transport failed: reason 'incomplete': retrying command
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk okay
>sd0: disk okay
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk not responding to selection
>sd0: disk okay
>The 2 things I see straight off are:
>
>1) /: out of inodes
>2) panic: error in swapping in u-area
> syncing file systems ...
> zs3: silo overflow
>
>I am not sure what exactly the culprit is here. I suspect it is a problem
>with swap however. My question is will the simply adding more swap via:
>
>sun1# mkfile 20m swapfile
>sun1# vi /etc/fstab
>sun1# swapon -a
>
>solve my problem? If not what do your eagle eyes spot that I am missing.

 The problem is apparently cabling. As was pointed out by several posters,
 the Mini-Din connections can hang quite loose at times, combine this with
 a myriad of cables hanging in the vicinity and you have potential trouble.

 I noticed the Mini-Din connection on the back of the IPX's CPU was a bit
 loose and part of the cause was a heavy monitor cable perched right on top.

 According to Sun-Support < Yes we *DO* have a contract :-), it just took me
 several weeks to discover this due to some confusion here at the office and
 some confustion at local Sun Office >, the checklist for this error should
 be as follows:

 1) Check All SCSI Cabling
 2) Check all power cables
 3) Check internal power cables on the disk
 4) When swapping cables, it is best to work with equipment that you *KNOW* is
    functional < not something suspect >.

 Thus, In Summary this problem is not the result of a disk error. If it were
 Apparently we would have received a number of hard errors.

 Many Thanks to EVERYONE who responded... Especially for steering me away from
 thinking it was a swap error.

 
 I *HUMBLY* Bow before the GREAT SUN gods

 Scott Surguine



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:28 CDT