[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference turris::digital_unix

Title:DIGITAL UNIX(FORMERLY KNOWN AS DEC OSF/1)
Notice:Welcome to the Digital UNIX Conference
Moderator:SMURF::DENHAM
Created:Thu Mar 16 1995
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:10068
Total number of notes:35879

8630.0. "Async I/O changes in 4.0?" by NETRIX::"[email protected]" (Steve Hancock) Wed Jan 29 1997 09:27

Hi folks,

We have been seeing a number of problems with customers
using Sybase 10 and 11 on Digital UNIX 4.0, 4.0a and 4.0b
that are getting corruptions in the database. Most of these
customers were happily running under 3.2x or 3.0 before
upgrading to 4.0x, but the corruption started after. No
amount of upgrading or installng patches has helped. 

I have discussed this problem with Sybase technical support
and they are of the opinion that changes were made in DU
4.0 in the Async. I/O area that is causing this problem.
Can someone knowledgable in that area comment on this? We
need to get a resolution to this problem as soon as possible.


Regards,

Steve Hancock
Digital UNIX Kernel Support
Atlanta CSC

[Posted by WWW Notes gateway]
T.RTitleUserPersonal
Name
DateLines
8630.1MBALDY::BRUCEour middle name is 'Equipment'Thu Jan 30 1997 11:137
You probably know this, but there *is* a patch for Oracle 7.3.2 for use on 
UNIX 4.0a which Oracle released to fix a bug in their software exposed by an
enhancement, I believe, to async i/o in 4.0.

Maybe Sybase needs to make a similar patch available....

Bruce Langston
8630.2vm: new-wire-method = 0SAPEC3::WALLMEROTHMon Feb 03 1997 17:156
    try to set vm:new'vm: new-wire-method = 0' in /etc/sysconfigtab.
    We encountered corruptions with INFORMIX and ADABAS. Setting
    'vm: new-wire-method = 0' solved the problem for ADABAS and
    DUNIX V4.0/V4.0A. I will test V4.0B this week.
    INFORMIX still has problems.
    See also notes 7828 and 8282.
8630.3status ?SAPEC3::WALLMEROTHThu Feb 20 1997 05:296
    re .0
    
    Can you please give a status of your problem.
    Has it been solved ?   How ?
    
    Thanks, Eckhardt