home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.ultrix
- Path: sparky!uunet!decwrl!deccrl!news.crl.dec.com!rdg.dec.com!decvax.dec.com!decvax.DEC.COM!jag
- From: jag@decvax.DEC.COM (John A. Gallant UEG)
- Subject: Re: SCSI/CAM problems
- Message-ID: <1992Sep8.171248.21775@decvax.dec.com>
- Sender: usenet@decvax.dec.com (Usenet News System)
- Nntp-Posting-Host: witsend.zk3.dec.com
- Reply-To: jag@zk3.dec.com
- Organization: OSF Engineering, Digital Equipment Corp.
- References: <ROBM.92Sep4150116@ataraxia.Berkeley.EDU> <1992Sep5.011659.7675@news.iastate.edu> <Bu862I.Dq@ie.utoronto.ca>
- Date: Tue, 8 Sep 1992 17:12:48 GMT
- Lines: 49
-
- In article <Bu862I.Dq@ie.utoronto.ca>, andy@ie.utoronto.ca (Andy Sun) writes:
- >john@iastate.edu (John Hascall) writes:
-
- >>What happens is we start getting *tons* of:
- >> XPT Packet Pool HIGH Water Mark Reached.
- >> cam_logger: CAM_ERROR packet
- >> cam_logger: No associated bus target lun
- >>messages on the console and also (in the two times I have seen it happen)
- >>the dreaded "cant get mbufs" message also appears. It seems to be somehow
- >>related to uptime (memory leak?) and load (so, of course, today, with a
- >>machine which hung last night and being the Friday before a long weekend,
- >>we had neither and I couldn't reproduce it for DEC *sigh*).
- >
- >I am afraid that the same thing happens to us (a DECsystem 5000/200
- >recently upgraded to 4.2a with SCSI/CAM). Similar messages appear in the
- >error log (viewed through uerf). ......................................
-
- How similar ?, are they High / Low or both water marks ? Please try
- the /usr/etc/cam_report program. The uerf utility is not able to decode
- all of the CAM error log information. We had to provide a second log
- decoder that provides more than a person could dream of.
-
- >................................ Even more horrible, our machine hung up
- >on us twice so far (i.e. no response, not even from console). I attempted
- >a core trace without any luck. It seems to be related to our USENET news
- >activities, when there were lots of read/write to an RZ58 partition.
-
- Try increasing the pool size ? What does the vmstat -K on the
- cores look like ? A large buffer cache in the file system code can
- result in a big "spike" in disk I/O when a flush occurs.
-
- >>I was told that backing out of SCSI/CAM was NOT as simple as
- >>just "setld -d ..." If we can't get a fix by next week we are
- >>resigned to going backwards to 4.2a.
- >
- >OH MY GOD!! (fainted...)
-
- No No No (waving ammonia salts ...), only for the CAMBASE* subset is
- there a problem. For the kernel, a delete and a rebuild is all that is
- needed.
-
- --
- John A. Gallant jag@zk3.dec.com
- Software Engineer - OSF Engineering Group
- Digital Equipment Corp. (603) 881-2472
-
- In the common people there is no wisdom, no penetration, no
- power of judgment.
- Marcus Cicero
-