ANK@CUNYVMS1.BITNET (04/22/87)
Since the past two-three days, we have been having problems on
one of our user-disks
Users get errors doing any file editing related tasks
The message is of the form
ACP-I-CRE file create failed
SYSTEM-HEADERFULL file header full
The manual on System messages and also the RMS file application
messages are vague on this... ??? We are running VMS 4.4 on 11/780
Also SHO DEV DR gives sufficient space free on all disks including
the SYSTEM disk .
Any suggestions as to what is going wrong ? and what to do to prevent
it. ?
Anil
BITNET:ank@cunyvms1
INTERNET:ank%cunyvms1.BITNET@wiscvm.educarl@CITHEX.CALTECH.EDU (Carl J Lydick) (04/24/87)
> Users get errors doing any file editing related tasks > The message is of the form > ACP-I-CRE file create failed > SYSTEM-HEADERFULL file header full > The manual on System messages and also the RMS file application > messages are vague on this... ??? We are running VMS 4.4 on 11/780 > Also SHO DEV DR gives sufficient space free on all disks including > the SYSTEM disk . The SYSTEM MESSAGES AND ERROR RECOVERY PROCEDURES manual isn't exactly vague on this; it merely assumes that you have some knowledge of the FILES-11 ON-DISK-STRUCTURE (LEVEL 2) that isn't normally available. The info in the SMAERP manual is: The file header map area on the volume is full.... This means that you've exceeded the Maximum files allowed value for the volume (actually, although it is referred to this way, it is actually the maximum file headers allowed, which is the upper limit of maximum files allowed). This can occur for several reasons: 1) You've got a highly fragmented disk, so that every file has lots of file headers associated with it. In this case the recommended action (compress the disk by doing an image backup and restore) will fix the problem. 2) You may have lots of little files. In this case, the backup and restore won't fix the problem. Instead, what you have to do is: A) Do an image backup of the disk. B) Initialize the disk with a larger value for max files. C) Restore the backed up files using BACKUP/NOINIT 3) The volume may be corrupted. Lots of luck, you'll need it. Hope this helps a little.
LEICHTER-JERRY@YALE.ARPA (04/26/87)
Since the past two-three days, we have been having problems on
one of our user-disks
Users get errors doing any file editing related tasks
The message is of the form
ACP-I-CRE file create failed
SYSTEM-HEADERFULL file header full
The manual on System messages and also the RMS file application
messages are vague on this... ??? We are running VMS 4.4 on 11/780
Also SHO DEV DR gives sufficient space free on all disks including
the SYSTEM disk .
The System Messages book (V4.2, the most recent I have around) has the
following to say:
Explanation: The file header map area ... is full and cannot be
extended. This error occurs only when the file header
extension is inhibited, for example, when the index file
is extended.
Rather terse - you have to know how ODS 2, the VMS file system, works to make
any sense of it at all. Basically, file headers are stored in the index file,
and the index file bitmap keeps track of which entries in the index file are
in use. The index file includes an entry for itself - for the most part, it
looks like a normal file - and the index file bitmap, too. Either of these
can, under normal conditions, be extended. This error apparently occurs when
both have to be extended. I don't claim to understand the details, but from
the text of the error message - and from personal experience - I'd say this is
a very rare occurance.
The System Messages book goes on to say:
User Action: Compress the volume by copying it with BACKUP.
That should fix the problem.
As a temporary work-around, deleting a bunch of files will free up some file
headers, which could then most likely be re-used. (Since I've never seen
this error, I can't say for certain - it could be that a disk in this state
just remains unable to accept new files even if headers are freed up. But
it's worth a try if re-building the disk RIGHT NOW would be a big headache.)
-- Jerry
-------SYSTEM@CRNLNS.BITNET (04/26/87)
Anil, As I understand it, the HEADERFULL message is a known problem with VMS 4.5 and earlier. It may or may not be fixed in 4.6. It is telling you that it can't extend the index file on the volume for some reason. I think one cause may be that the disk structure is too badly fragmented. At any rate, you will have to do a full Backup and restore of the disk. My solution is to initialize our user disk with a preallocated index file that is large enough for all the files that will be created. Also, you must not let a Backup "image mode restore" initialize the disks itself. (Specify /NOINITIALIZE to Backup when you start the restore.) This also gets around the problem that some sites have seen where two cluster members try to extend the index file simultaneously and both manage to do it, creating two incompatible extensions. This bug is also reported to be under investigation. Our user disks typically have about 30,000 files on each. The following command preallocates 40,000 file headers: $ INIT HSC000$DUA5: SCRATCH_DISK /HEADERS=40000/CLUSTER=8/EXTENSION=32 I hope this helps. Selden E. Ball, Jr. (Wilson Lab's network and system manager) Cornell University NYNEX: +1-607-255-0688 Laboratory of Nuclear Studies BITNET: SYSTEM@CRNLNS Wilson Synchrotron Lab ARPA: SYSTEM%CRNLNS.BITNET@WISCVM.WISC.EDU Judd Falls & Dryden Road PHYSnet/HEPnet/SPAN: Ithaca, NY, USA 14853 LNS61::SYSTEM = 44283::SYSTEM (node 43.251)