(A1) What to do if disk space is low on the system disk for VMS
This applies to all VMS system disks (Felix, MVG etc.)
(a) log-in as system
Determine name of system disk as follows:
$ show log sys$sysdevice
(b) Purge the error log files:
$ set def dua0:[sys0.syserr] (if dua0: is system disk)
$ purge errlog.sys
if that doesn't free up enough space then,
$ delete errlog.sys;*
(c) Also, do:
$ purge sys$manager:*.*
$ purge sys$system:*.*
to free additional space.
(d) Also purge the accounting file as follows:
$ set account/new_file
$ purge sys$manager:accountng.dat/log
Back to Top
(A2) What to do if system disk out of space, and VMS won't boot up
If the system won't boot because of zero free blocks on the system disk, then the first step is to perform a "conversational boot", which gets you to the SYSBOOT prompt and lets you run DCL commands to free up some space. The way to start a conversational boot varies with the specific type or Vaxstation or MicroVAX. For an Alphastation 200 or 400 (ME2, MVL), halt the machine and type the following: >>> BOOT -FLAGS 0,1 dka0 This should take you a the SYSBOOT> prompt, from where you can try to free up space as follows: SYSBOOT> SET/STARTUP OPA0: SYSBOOT> CONTINUE This brings up the DCL prompt, type commands to free up space, e.g.: $ set def sys$sysroot:[sysexe] $ purge *.*/log $ set def sys$sysroot:[sysmgr] $ purge *.*/log $ purge dka0:[000000...]*.*/log etc. This should free up enough space to allow a normal boot. Proceed as follows: $ RUN SYS$SYSTEM:SYSGEN $ SET/STARTUP SYS$SYSTEM:STARTUP.COM $ WRITE CURRENT $ EXIT $ SYS$SYSTEM:STARTUP This should restart the system normally.Back to Top
Examples of user disks that may run low on space are MVGD: MVFD: MVCD:
and MEID: (To know the actual disk name, use "$ show log MVGD" etc.)
(1) Log-in as system and check free space with "show dev dk" and
"show dev du"
(2) If space is low, first ask anyone who has big files on that disk
whether they can be deleted or moved elsewhere (disk/tape etc.)
(3) Delete all temp. plot files as follows:
(assuming MVGD: is our nearly full disk)
(Caution: a typo in the following command could be disastrous!)
$ delete mvgd:[000000...]pl*.tmp;*/log
(4) As a last resort, do a disk-wide purge:
$ purge mvgd:[000000...]*.*/log
Back to Top
(a) log-in as system
(b) $ set def dua0:[sys0.syserr]
$ purge errlog.sys
if that doesn't free up enough space then,
$ delete errlog.sys;*
(c) Also, do:
$ purge sys$manager:*.*
$ purge sys$system:*.*
to free additional space.
Back to Topkochhar@physiology.wisc.edu