| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 172.1 | | LEFTY::CWILLIAMS | CD or not CD, that's the question | Thu Jul 28 1994 09:06 | 5 |
| 172.2 | | CSC32::D_STUART | Keep it clean, close and loaded | Thu Jul 28 1994 11:37 | 11 |
| 172.3 | | LEFTY::CWILLIAMS | CD or not CD, that's the question | Thu Jul 28 1994 13:30 | 6 |
| 172.4 | How to correct MOUNT-F-OBSOLETE | BABAGI::SLEDD | | Tue Aug 02 1994 16:28 | 11 |
| 172.5 | RMS%....INDEX FILE FULL | KERNEL::CLARK | STRUGGLING AGAINST GRAVITY... | Thu Mar 20 1997 07:58 | 32 |
| Along similar lines...
Customer has RW530 Jukebox with 60 mounted JBA WORM platters RWX5K-02
(1.2Gbyte-512 bytes/sector).
He has run this config for over a year, using the command:-
INIT/MAX=3500 JBA%: MANddmmmyy to initialise his volumes.
He has found recently that when approximately 1900 files have been
written to a JBA volume, he gets error messages:-
ACP FILE CREATE FAILED
%RMS "INDEX FILE FULL"
I haven't found the "INDEX FILE FULL" message in any of JKCONFIG,
JBUTIL or ODUTIL message tables.
The nearest I found was NOROOMFORHDR "no room in index area for new
headers" which has a similar meaning.
I am concerned that it was RMS which reported the condition. Can
anybody shed some more light on this particular issue please? Is it an
RMS condition or threshold which is causing this problem or is RMS
reporting a problem caused by the WORM limitations?
The customer has not identified the cluster size in use for this
media. Could it be that indexf.sys is unable to deal with the number of
file headers?
Dave Clark
|
| 172.6 | from W11AACP.EXE | TAPE::SENEKER | Head banging causes brain mush | Thu Mar 20 1997 12:57 | 22 |
| Dave,
The WORMS-11 ACP (W11AACP.EXE) returns this status when the index
file is full.
A possible reason for this happening at 1900 files is that the
directory file headers are being modified multiple times as the
1900 files are created. Please note, on a WORMS-11 disk each time
a file is modified and written back to the disk a new file header
is created in the index file. You can use ODUTIL to verify the
state of the index file.
A message help file should exist in SYS$HELP:PERCEPTICS.MSGHLP$DATA
that contains the OSMS meaning of various error messages. Of the
top of my head I don't remember how to get it to be included in the
list that the DCL command HELP/MESSAGE searches through. You should
be able to do a $ HELP/MESSAGE IDXFILEFULL and it would tell you
how/why that message is generated.
The cluster size is fixed for WORMS-11 media and is not an issue.
Rob
|