USAGE for mkdirs.sh v4.7.0:
mkdirs.sh <instance> [-s <ServerID>] [-t <server_type>] [-I <svc>[,<svc2>]] [-MDD /bigdisk] [-MLG /jnl] [-MDB1 /db1] [-MDB2 /db2] [-f] [-p] [-test [-clean]] [-n] [-L <log>] [-d|-D]
This script initializes an SDP instance on a single machine.
This script is intended to support two scenarios:
* First time SDP installation on a given machine.
* Adding new SDP instances (separate Helix Core data sets) to an existing
SDP installation on a given machine.
And SDP instance is a single Helix Core data set, with its own unique
set of one set of users, changelist numbers, jobs, labels, versioned
files, etc. An organization may run a single instance or multiple
This is intended to be run either as root or as the operating system
user account (OSUSER) that p4d is configured to run as, typically
'perforce'. It should be run as root for the initial install.
Subsequent additions of new instances do not require root.
If an initial install as done by a user other than root, various
directories must exist and be writable and owned by 'perforce' before starting:
This script creates an init script in the /p4/N/bin directory.
After running this script, set up the crontab based on templates
generated in /p4/common/etc/cron.d. For convenience, a sample cronat
is generated for the current machine in /p4/common/etc/cron.d named
where <osuser> is the user that services run as (typically 'perforce'),
and <host> is the short hostname (as returned by a 'hostname -s' command).
Next, put the license file in place in the P4ROOT dir, and launch the server
with the init script.
Then run /p4/common/bin/p4master_run instance /p4/common/bin/live_checkpoint.sh
and then run both the daily_checkpoint.sh and recreate_db_checkpoint.sh to
make sure everything is working before setting up the crontab.
Also run /p4/common/bin/p4master_run <instance> /p4/common/bin/p4review.py <instance>
to make sure the review script is working properly. If you intend to use
Swarm, you can skip configuration of the review daemon, and instead configure
Swarm to handle review-style email notifications.
Specify the SDP instance name to add. This is a reference to the Perforce
Helix Core data set.
Specify the ServerID, overriding the REPLICA_ID setting in the configuration
Specify the ServerID of the P4TARGET of the server being installed.
Use this only when setting up an HA replica of an edge server.
Specify the server type, overriding the SERVER_TYPE setting in the config
file. Valid values are:
* p4d_master - A master/commit server.
* p4d_replica - A replica with all metadata from the master (not
filtered in any way).
* p4d_filtered_replica - A filtered replica or filtered forwarding
* p4d_edge - An edge server.
* p4d_edge_replica - Replica of an edge server. If used,
'-S <TargetServerID>' is required.
* p4broker - An SDP host running only a broker, with no p4d.
* p4proxy - An SDP host running a proxy (maybe with a broker in front),
with no p4d.
Specify additional init scripts to be added to /p4/<instance>/bin
for the instance.
By default, the p4p service is installed only if '-t p4proxy' is
specified, and p4dtg is never installed by default. Valid values
to specify are 'p4p' and 'dtg' (for the P4DTG init script).
If services are not installed by default, they can be added later
using templates in /p4/common/etc/init.d. Also, templates for
systemd service files are supplied in /p4/common/etc/systemd/system.
Specify the '-M*' to specify mount points, overriding DD/LG/DB1/DB2
settings in the config file. Sample:
-MDD /bigdisk -MLG /jnl -MDB1 /fast
If -MDB2 is not specified, it is set the the same value as -MDB1 if
that is set, or else it defaults to the same default value as DB1.
-f Specify -f 'fast mode' to skip chown/chmod commands on depot files.
This should only be used when you are certain the ownership and
permissions are correct, and if you have large amounts of existing
data for which the chown/chmod of the directory tree would be
-p Specify '-p' to halt processing after preflight checks are complete,
and before actual processing starts. By default, processing starts
immediately upon successful completion of preflight checks.
Specify the path to a log file, or the special value 'off' to disable
logging. By default, all output (stdout and stderr) goes to this file
in the current directory:
NOTE: This script is self-logging. That is, output displayed on the
screen is simultaneously captured in the log file. Do not run this
script with redirection operators like '> log' or '2>&1', and do not
Specify '-test' to execute a simulated install to /tmp/p4 as the install
root (rather than /p4), and with the mount point directories specified in
the configuration file prefixed with /tmp/hxmounts, defaulting to:
Specify '-clean' with '-test' to clean up from prior test installs,
which will result in removal of files/folders installed under /tmp/hxmounts
Do not specify '-clean' if you want to test a series of installs.
-n No-Op. In No-Op mode, no actions that affect data or structures are
taken. Instead, commands that would be run are displayed. This is
an alternative to -test. Unlike '-p' which stops after the preflight
checks, with '-n' more processing logic can be exercised, with greater
detail about what commands that would be executed without '-n'.
-d Increase verbosity for debugging.
-D Set extreme debugging verbosity, using bash '-x' mode. Also implies -d.
-h Display short help message
-man Display man-style help message
The mkdirs.sh script uses a configuration file for many settings. A
sample file, mkdirs.cfg, is included with the SDP. After determining
your SDP instance name (e.g. '1' or 'abc'), create a configuration
file for it named mkdirs.<N>.cfg, replacing 'N' with your instance.
Running 'mkdirs.sh N' will load configuration settings from mkdirs.N.cfg.
This script can be useful in testing and upgrading to new versions of
the SDP, when the '-test' flag is used.
Example 1: Setup of first instance
Setup of the first instance on a machine using the default instance name,
'1', executed after using sudo to become root:
$ sudo su -
$ cd /hxdepots/sdp/Server/Unix/setup
$ vi mkdirs.cfg
# Adjust settings as desired, e.g P4PORT, P4BROKERPORT, etc.
$ ./mkdirs.sh 1
A log will be generated, mkdirs.1.<timestamp>.log
Example 2: Setup of additional instance named 'abc'.
Setup a second instance on the machine, which will be a separate Helix
Core instance with its own P4ROOT, its own set of users and
changelists, and its own license file (copied from the master instance).
Note that while the first run of mkdirs.sh on a given machine should be
done as root, but subsequent instance additions should be done as the
'perforce' user (or whatever operating system user accounts Perforce
Helix services run as).
$ sudo su - perforce
$ cd /hxdepots/sdp/Server/Unix/setup
$ cp -p mkdirs.cfg mkdirs.abc.cfg
$ vi mkdirs.abc.cfg
# Adjust settings in mkdirs.abc.cfg as desired, e.g P4PORT, P4BROKERPORT, etc.
$ ./mkdirs.sh abc
A log will be generated, mkdirs.abc.<timestamp>.log
Example 3: Setup of additional instance named 'alpha' to run a p4p:
$ ./mkdirs.sh alpha -t p4proxy