lockfile - conditional semaphore-file creator
lockfile -sleeptime | -r retries |
-l locktimeout | -s suspend | -! | -ml | -mu | filename ...
lockfile can be used to create one or more semaphore files. If lock-
file can’t create all the specified files (in the specified order), it
waits sleeptime (defaults to 8) seconds and retries the last file that
didn’t succeed. You can specify the number of retries to do until
failure is returned. If the number of retries is -1 (default, i.e.,
-r-1) lockfile will retry forever.
If the number of retries expires before all files have been created,
lockfile returns failure and removes all the files it created up till
Using lockfile as the condition of a loop in a shell script can be done
easily by using the -! flag to invert the exit status. To prevent
infinite loops, failures for any reason other than the lockfile already
existing are not inverted to success but rather are still returned as
All flags can be specified anywhere on the command line, they will be
processed when encountered. The command line is simply parsed from
left to right.
All files created by lockfile will be read-only, and therefore will
have to be removed with rm -f.
If you specify a locktimeout then a lockfile will be removed by force
after locktimeout seconds have passed since the lockfile was last modi-
fied/created (most likely by some other program that unexpectedly died
a long time ago, and hence could not clean up any leftover lockfiles).
Lockfile is clock skew immune. After a lockfile has been removed by
force, a suspension of suspend seconds (defaults to 16) is taken into
account, in order to prevent the inadvertent immediate removal of any
newly created lockfile by another program (compare SUSPEND in proc-
If the permissions on the system mail spool directory allow it, or if
lockfile is suitably setgid, it will be able to lock and unlock your
system mailbox by using the options -ml and -mu respectively.
Suppose you want to make sure that access to the file "important" is
serialised, i.e., no more than one program or shell script should be
allowed to access it. For simplicity’s sake, let’s suppose that it is
a shell script. In this case you could solve it like this:
rm -f important.lock
Now if all the scripts that access "important" follow this guideline,
you will be assured that at most one script will be executing between
the ‘lockfile’ and the ‘rm’ commands.
LOGNAME used as a hint to determine the invoker’s login-
/etc/passwd to verify and/or correct the invoker’s loginname
(and to find out his HOME directory, if needed)
lockfile for the system mailbox, the environment
variables present in here will not be taken from
the environment, but will be determined by look-
ing in /etc/passwd
rm(1), mail(1), binmail(1), sendmail(8), procmail(1)
Filename too long, ... Use shorter filenames.
Forced unlock denied on "x"
No write permission in the directory where lock-
file "x" resides, or more than one lockfile try-
ing to force a lock at exactly the same time.
Forcing lock on "x" Lockfile "x" is going to be removed by force be-
cause of a timeout (compare LOCKTIMEOUT in proc-
Out of memory, ... The system is out of swap space.
Signal received, ... Lockfile will remove anything it created till
now and terminate.
Sorry, ... The retries limit has been reached.
Truncating "x" and retrying lock
"x" does not seem to be a valid filename.
Try praying, ... Missing subdirectories or insufficient privi-
Definitely less than one.
The behavior of the -! flag, while useful, is not necessarily intu-
itive or consistent. When testing lockfile’s return value, shell
script writers should consider carefully whether they want to use the
-! flag, simply reverse the test, or do a switch on the exact exit-
code. In general, the -! flag should only be used when lockfile is
the conditional of a loop.
Lockfile is NFS-resistant and eight-bit clean.
Calling up lockfile with the -h or -? options will cause it to display
a command-line help page. Calling it up with the -v option will cause
it to display its version information.
Multiple -! flags will toggle the return status.
Since flags can occur anywhere on the command line, any filename start-
ing with a ’-’ has to be preceded by ’./’.
The number of retries will not be reset when any following file is be-
ing created (i.e., they are simply used up). It can, however, be reset
by specifying -rnewretries after every file on the command line.
Although files with any name can be used as lockfiles, it is common
practice to use the extension ‘.lock’ to lock mailfolders (it is ap-
pended to the mailfolder name). In case one does not want to have to
worry about too long filenames and does not have to conform to any oth-
er lockfilename convention, then an excellent way to generate a lock-
filename corresponding to some already existing file is by taking the
prefix ‘lock.’ and appending the i-node number of the file which is to
This program is part of the procmail mail-processing-package (v3.22)
available at http://www.procmail.org/ or ftp.procmail.org in pub/proc-
There exists a mailinglist for questions relating to any program in the
for submitting questions/answers.
for subscription requests.
If you would like to stay informed about new versions and official
patches send a subscription request to
(this is a readonly list).
Stephen R. van den Berg
Philip A. Guenther
BuGless 2001/06/23 LOCKFILE(1)
Man(1) output converted with