home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!sdd.hp.com!hplabs!ucbvax!CLUSTER.POLY-NORTH-LONDON.AC.UK!SYSTEM
- From: SYSTEM@CLUSTER.POLY-NORTH-LONDON.AC.UK (System Manager)
- Newsgroups: comp.os.vms
- Subject: HEADERFULL messages when copying 2xRA81 to 1xRF72.
- Message-ID: <9209041047.AA08340@ucbvax.Berkeley.EDU>
- Date: 3 Sep 92 17:15:00 GMT
- Sender: daemon@ucbvax.BERKELEY.EDU
- Distribution: world
- Organization: The Internet
- Lines: 170
-
- This posting is 169 lines long.
-
- Problems once again! We (We? I mean muggins here) are in the middle of
- swapping our old CI based 8000 cluster with RA81 disks for a new 4500 (x4)
- RF72 LAVC VMS 5.5-1 cluster. I've got a problem copying some disks with
- HEADERFULL messages (detailed below), and basically need to know how to
- locate the headers used / headers free values on a disk volume.
-
- Here follows a more detailed synopsis/diary of the problem ...
-
-
- Step 1. Initialize the disk with the command INIT $1$DIA11: DISK11
-
- Step 2. Back off my old RA81's to my new RF72. I excluded all the top level
- .SYS files in case they interfered with the initialize command
- already executed on the device.
-
- $ set noon
- $ mount/NoRebuild $1$dia11 disk11
- %MOUNT-I-MOUNTED, DISK11 mounted on _$1$DIA11: (IDUN1)
- $ backup/new_version/ign=(label,inter)/Exclude=[000000]*.sys -
- USERSA:[*...]*.*;* -
- $1$dia11:[*...]*.*;*/Owner=Original
- $ dismo $1$dia11
- $ set noon
- $ mount/NoRebuild $1$dia11 disk11
- %MOUNT-I-MOUNTED, DISK11 mounted on _$1$DIA11: (IDUN1)
- $! mount/norebuild 'p2 'p2
- $ backup/new_version/ign=(label,inter)/Exclude=[000000]*.sys -
- USERSC:[*...]*.*;* -
- $1$dia11:[*...]*.*;*/Owner=Original
- %BACKUP-E-OPENOUT, error opening $1$DIA11:[77SULLIVAN]E.LIS;9 as output
- -RMS-E-CRE, ACP file create failed
- -SYSTEM-W-HEADERFULL, file header is full
- %BACKUP-E-OPENOUT, error opening $1$DIA11:[77SULLIVAN]E.SUL;4 as output
- -RMS-E-CRE, ACP file create failed
- -SYSTEM-W-HEADERFULL, file header is full
- %BACKUP-E-OPENOUT, error opening $1$DIA11:[77SULLIVAN]EHO.DAT;1 as output
- -RMS-E-CRE, ACP file create failed
- -SYSTEM-W-HEADERFULL, file header is full
- . . .
- . . .
- . . . Etc.
-
- Hmmm. The documentation seems to suggest that the default values are
- usually sufficient. My RA81 hasn't been initialised in yonks, so I don't
- know if whoever did it put in special /MAXIMUM_FILES or /HEADERS
- qualifiers. I think not, as looking at the INDEXF.SYS files on the top
- level [000000] of these RA81's gets me the following ...
-
- Directory USERSA:[000000]
-
- INDEXF.SYS;1 69814/69816 20-FEB-1987 15:51:40.94
-
- Total of 1 file, 69814/69816 blocks.
-
- Directory USERSC:[000000]
-
- INDEXF.SYS;1 26107/26109 20-APR-1988 14:12:20.50
-
- Total of 1 file, 26107/26109 blocks.
-
- ... There doesn't seem to be any pre-extension of INDEXF.SYS here.
- My RF72 (DIA11) has a INDEXF.SYS size of 49188/49188 when it bombs out ,
- which is less than the INDEXF.SYS size on USERSA. Why does my RA81 handle
- a larger INDEXF.SYS size than my RF72?
-
- Step 3. Check the fragmentation on my RA81's. Looks fine.
- We run DISKIT from UIS and haven't any complaints about it. The
- FRAG report says the disks are in pretty reasonable shape ...
-
- Step 4. Check the RF72's statistics ...
-
- Disk $1$DIA11: (IDUN1), device type RF72, is online, mounted, file-oriented
- device, shareable, available to cluster, error logging is enabled.
-
- Error count 0 Operations completed 329590
- Owner process "" Owner UIC [SYSTEM]
- Owner process ID 00000000 Dev Prot S:RWED,O:RWED,G:RWED,W:RWED
- Reference count 1 Default buffer size 512
- Total blocks 1954050 Sectors per track 50
- Total cylinders 1861 Tracks per cylinder 21
- Host name "IDUN1" Host type, avail VAX 4000-500, yes
- Allocation class 1
-
- Volume label "DISK11" Relative volume number 0
- Cluster size 3 Transaction count 1
- Free blocks 709050 Maximum files allowed 244256
- Extend quantity 5 Mount count 5
- Mount status System Cache name "_HSC000$DUA10:XQPCACHE"
- Extent cache size 64 Maximum blocks in extent cache 70905
- File ID cache size 64 Blocks currently in extent cache 0
- Quota cache size 0 Maximum buffers in FCP cache 1246
-
- Volume status: subject to mount verification, file high-water marking, write-
- through caching enabled.
- Volume is also mounted on MAIA, LOFN, IDUN, LOFN1.
-
- DIR $1$DIA11:[000000...]/Grand_Total/Size=All
-
- GRAND TOTAL OF 1361 DIRECTORIES, 49116 files, 1179190/1245000 blocks.
-
- <SHOW DEVICE/FULL on one of my RA81's here for comparison>
-
- Disk HSC000$DUA16:, device type RA81, is online, mounted, file-oriented device,
- shareable, available to cluster, error logging is enabled.
-
- Error count 0 Operations completed 25150
- Owner process "" Owner UIC [SYSTEM]
- Owner process ID 00000000 Dev Prot S:RWED,O:RWED,G:RWED,W:RWED
- Reference count 15 Default buffer size 512
- Total blocks 891072 Sectors per track 51
- Total cylinders 1248 Tracks per cylinder 14
- Host name "HSC000" Host type, avail HS50, yes
-
- Volume label "USERSC" Relative volume number 0
- Cluster size 3 Transaction count 15
- Free blocks 104598 Maximum files allowed 111384
- Extend quantity 5 Mount count 4
- Mount status System Cache name "_HSC000$DUA10:XQPCACHE"
- Extent cache size 64 Maximum blocks in extent cache 10459
- File ID cache size 64 Blocks currently in extent cache 2460
- Quota cache size 164 Maximum buffers in FCP cache 1246
-
- Volume status: subject to mount verification, write-through caching enabled.
- Volume is also mounted on LOFN, MAIA, IDUN.
-
-
- Step 5. Check to see what happens when I INIT a disk with, say, 240,000
- headers. Yikes!! It preallocates the space in my INDEXF.SYS file!
- Can't afford that much space wasted. That's approx. 25% of my disk!
-
- HHmmm. What's the point in having 244,000 allowable files when the headers
- run out? (at 20% of the allowable files value too ). More worrying from my
- point of view, is that the disks that have been backed up successfully might
- have only one header left free and I wouldn't know about it!
-
- O.K., If I have to tell the disk how many headers to expect I want to be
- able to make the best guess possible based on current usage. I will also
- want to keep an eye on it as the disk starts to fill up to make sure that I
- can Initialize it again over a weekend with a greater /HEADERS value when I
- get forewarned about this so as not to upset the users.
-
- Question : How can I find out how many file headers I have USED/FREE on a
- disk?
-
- Postscript : I have been talking to DEC CSC in the UK and they basically
- said 'tough luck - guess the headers and do a backup/init/restore when you
- think you may have a problem'. Mind you, this was after getting me a patch
- that they thought would fix it - F11X$01_U1055. I applied that and it
- didn't make any difference.
-
-
- Any suggestions, requests for more info', or good whisky recommendations
- gratefully received.
-
- Regards,
-
- Brian.
- ------------------------------------------------------------------------------
- | | |\ | | Post: The University Of North London
- | | | \ | | 166-220 Holloway Road, London, N7 8DB
- | | | \ | | Voice: +44 71 700 4900 Janet: Brian@UK.AC.PNL.CLSTR
- \---/ | \| |_____ Fax : +44 71 753 5120 DTE : 000045401000
- Disclaimer: I said WHAT??? Flames : NL:
- It's 106 miles to Chicago, we've got a full tank of gas, half a packet of
- cigarettes, it's dark and we're wearing sunglasses... Hit It! (Blues Brothers)
- ------------------------------------------------------------------------------
-
-
-