TIP: Click on subject to list as thread! ANSI
echo: lan
to: LEE ARONER
from: MIKE BILOW
date: 1997-12-22 03:29:00
subject: NOVELL & WD 6.4GIG

LEE ARONER wrote in a message to MIKE BILOW:
 LA>    Jeez....you learn something every day in this echo!!
We try.  :-)
 LA>    I have never heard of this...
The importance of disabling hardware write cache on drives connected to 
NetWare servers has been thoroughly documented.  See, for example, Novell TID 
2914137, which discusses the issue in the context of Micropolis drives and 
gives specific instructions on how to disable their write caches, but makes 
clear that the need to disable write caches is applicable to all brands of 
drive.
 LA>    I wonder if this is the source of the drive corruption 
 LA> problems I'm seeing on my Seagate Hawk 2 gigers running under 
 LA> Warp Server? 
It's quite likely; Novell TID 2925087 specifically mentions the Seagate Hawk 
and its NetWare incompatibility with write cache enabled.  This TID is so 
on-target about your problem that I'll post an excerpt here:
   Novell has seen these types of errors on a regular basis with customers 
ho
   use large drives with the "write cache" enabled on them. Seagate's 
   Barracuda and Hawk drives will default the write cache on. Some Micropolis 
   drives also will enable this setting by default. The problem with a write 
   cache is that NetWare thinks it is writing to the physical drive when in 
   fact it is not. It finishes writing to the write cache of the drive and 
   then the write cache takes responsibility of moving the data to the 
   physical drive. Without going in depth on how NetWare writes are performed 
   through a write cache, the basic idea is that a write cache can lead to 
   data corruption and the above error messages. However, these error 
essages
   are not the only result of data corruption, there can be corruption on 
   files that are not compressed if the write cache is enabled. In other 
   words, data corruption can happen on all types of files, compressed or 
t.
I don't know if I can find a document from IBM in the OS/2 context on this, 
but I can assure you I have a lot of knowledge about how the internal SCSI 
stack and filesystem works in OS/2, and the discussion from Novell is equally 
applicable to OS/2 HPFS, more so for HPFS386.
 LA>    Where can I get more info on this?
You can get Seagate's official utility to manipulate the write cache enable 
bit on the appropriate device mode page, APSI-WCE, at:
   ftp://ftp.seagate.com/techsuppt/seagate_utils/aspiid15.zip
I can save you a lot of aggravation that you will encounter when trying to 
make sense of the Seagate docuemtation: for OS/2, turn off the write cache.
 
-- Mike
--- 
---------------
* Origin: N1BEE BBS +1 401 944 8498 V.34/V.FC/V.32bis/HST16.8 (1:323/107)

SOURCE: echomail via exec-pc

Email questions or comments to sysop@ipingthereforeiam.com
All parts of this website painstakingly hand-crafted in the U.S.A.!
IPTIA BBS/MUD/Terminal/Game Server List, © 2025 IPTIA Consulting™.