Skip to main content

A zc.buildout recipe for Plone deployments which configures various unix system services.

Project description

Introduction

This recipe provides support for configuring various unix services when deploying Plone/Zope2 with buildout.

As for now the following features are provided:

  • Create Logrotate configuration for all Zope 2 instance and ZEO parts, as well as ftw.structlog logfiles.

  • Create RedHat-like run-control scripts.

  • Create packall script for packing of all storages.

  • Create filebeat prospector configs for ftw.structlog and ftw.contentstats logs.

Supported options

The recipe supports the following options:

logrotate-directory

The directory where the logrotate configuration file will be created. Defaults to ${buildout:directory}/etc/logrotate.d. Add this parameter with no arguments to supress generation of logrotate configuration.

If this parameter is set, this recipe will create logrotate configs for all Zope 2 instance and ZEO parts that are present, and (unconditionally) a logrotate config for ftw.structlog logfiles.

logrotate-options

A list of logrotate options that should be added to the logrotate configuration.

Configuration for reopening rotated logfiles is added automatically if postrotate is not specified in logrotate-options.

Example:

logrotate-options =
    rotate 4
    weekly
    missingok
    notifempty
    nomail

The logrotate config for ftw.structlog logfiles will be created with settings similar to the other logfiles, except:

  • No postrotate script will be automatically inserted if not already present in logrotate-options

  • missingok will always be included

  • Rotation mode will always be copytruncate, and nocopytruncate will be ignored

startup-directory

If specified, a start script is created in the given directory. Generation of run-control scripts is disabled if this option is present.

shutdown-directory

If specified, a shutdown script is created in the given directory. Generation of run-control scripts is disabled if this option is present.

rc-directory

The directory where run-control scripts should be installed. Defaults to ${buildout:directory}/bin. Add this parameter with no arguments to supress generation of run-control scripts.

rc-prefix

Name prefix for run-control scripts. Defaults to rc-.

rc-user

The name of the user used to start processes by run-control scripts. Defaults to zope.

has-supervisor

Boolean indication whether supervisor is beeing used. If true, a run control script is generated for supervisor only. If false, run control scripts are generated for all zope and zeo parts. By default, the recipe tries to automatically determine if supervisor is beeing used.

zopes

A list of Zope 2 instance parts. Defaults to all parts using the plone.recipe.zope2instance recipe.

zeos

A list of ZEO server parts. Defaults to all parts using either the plone.recipe.zeoserver or plone.recipe.zope2zeoserver recipe.

packall-symlink-directory

Creates a symlink to the packall script in the given directory. Can be used to automate packing for multiple deployments.

create-filebeat-config

Boolean to indicate whether a filebeat config should be created for this deployment. Defaults to true. Set to false to suppress creation of a filebeat config.

Changelog

1.4.3 (2019-02-04)

  • Skip duplicate zope parts when generating logrotate config. [lgraf]

1.4.2 (2019-02-04)

  • Add option to suppress creation of filebeat config. [lgraf]

1.4.1 (2018-01-10)

  • Only chmod scripts to executable if they’re not already executable. [buchi]

1.4.0 (2017-12-08)

  • Generate filebeat prospector configs for ftw.structlog and ftw.contentstats logs. [lgraf]

  • Respect current umask when making scripts executable. [buchi]

1.3.0 (2017-09-11)

  • Generate logrotate config for ftw.structlog’s logfiles. [lgraf]

1.2.0 (2016-02-10)

  • Let bin/packall log on success to var/log/pack.log. [jone]

  • Add option to create startup and shutdown scripts in a global directory instead of run-control scripts. [buchi]

  • Create packall script for packing of all storages. [buchi]

  • Fix handling of logrotate options in buildout 2.x by no longer dropping the first option. [buchi]

1.1.1 (2014-01-21)

  • Make postrotate command fail silently if pid file is missing. [buchi]

1.1 (2013-03-05)

  • Added option to specify the user under which run-control scripts start processes. [buchi]

  • Added status command to supervisor run-control script. [buchi]

1.0.1 (2012-03-12)

  • Fixed handling of logrotate options containing whitespaces. [buchi]

1.0 (2012-03-12)

  • Initial release. [buchi]

Example usage

First we create a fake plone.recipe.zope2instance recipe, which we can use in our tests.

Create a recipes source directory:

>>> mkdir(sample_buildout, 'plone.recipe.zope2instance')

and then create a source file with the fake recipe:

>>> write(sample_buildout, 'plone.recipe.zope2instance',
...       'zope2instance.py',
... """
... import os, zc.buildout
...
... class Recipe(object):
...
...     def __init__(self, buildout, name, options):
...         self.name, self.options = name, options
...         options['event-log'] = os.path.join(
...                               buildout['buildout']['directory'],
...                               'var', 'log', self.name + '.log'
...                               )
...         options['z2-log'] = os.path.join(
...                               buildout['buildout']['directory'],
...                               'var', 'log', self.name + '-Z2.log'
...                               )
...
...     def install(self):
...         return tuple()
...
...     def update(self):
...         pass
... """)

Provide packaging information so that the recipe can be installed as a develop egg:

>>> write(sample_buildout, 'plone.recipe.zope2instance', 'setup.py',
... """
... from setuptools import setup
...
... setup(
...     name = "plone.recipe.zope2instance",
...     entry_points = {'zc.buildout': ['default = zope2instance:Recipe']},
...     )
... """)

Add a README.txt to avoid an annoying warning from distutils:

>>> write(sample_buildout, 'plone.recipe.zope2instance', 'README.txt', " ")

We’ll start by creating a simple buildout that uses our recipe:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance
... parts = instance1 deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Installing instance1.
Installing deployment.
<BLANKLINE>

We should now have a file with the same name as our buildout directory containing our logrotate configuration:

>>> cat(sample_buildout, 'etc', 'logrotate.d', 'sample-buildout')
... #doctest: -NORMALIZE_WHITESPACE
/sample-buildout/var/log/instance1.log
/sample-buildout/var/log/instance1-Z2.log {
    sharedscripts
    postrotate
        /bin/kill -SIGUSR2 `cat /sample-buildout/var/instance1.pid 2>/dev/null` >/dev/null 2>&1 || true
    endscript
}
/sample-buildout/var/log/instance1-json.log {
    copytruncate
    missingok
}

We should also have a run-control script for instance1:

>>> ls(sample_buildout, 'bin')
- buildout
- rc-instance1

>>> cat(sample_buildout, 'bin', 'rc-instance1')
#!/bin/sh
<BLANKLINE>
# chkconfig: 345 90 10
# description: Starts Zope
<BLANKLINE>
START_SCRIPT="/sample-buildout/bin/instance1"
<BLANKLINE>
[ -f $START_SCRIPT ] || exit 1
<BLANKLINE>
# Source function library.
. /etc/init.d/functions
<BLANKLINE>
RETVAL=0
<BLANKLINE>
if [ $(whoami) != "root" ]; then
    echo "You must be root."
    exit 1
fi
<BLANKLINE>
case $1 in
    start|stop)
        su zope -c "$START_SCRIPT $*" </dev/null
        RETVAL=$?
        if [ $RETVAL -eq 0 ]
        then
            echo_success
        else
            echo_failure
        fi
        echo
        ;;
    restart)
        ${0} stop
        sleep 1
        ${0} start
        ;;
    *)
        su zope -c "$START_SCRIPT $*" </dev/null
        ;;
esac

We should also have a filebeat prospectors config for your deployment:

>>> cat(sample_buildout, 'etc', 'filebeat.d', 'sample-buildout.yml')
... #doctest: -NORMALIZE_WHITESPACE
- type: log
  fields:
    event_type: contentstats
    deployment: sample-buildout
  fields_under_root: true
  json.keys_under_root: true
  json.add_error_key: true
  paths:
    - /sample-buildout/var/log/contentstats-json.log
- type: log
  fields:
    event_type: structlog
    deployment: sample-buildout
  fields_under_root: true
  json.keys_under_root: true
  json.add_error_key: true
  paths:
    - /sample-buildout/var/log/instance1-json.log

Except if we specifically disable creation of the filebeat config:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance
... parts = instance1 deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [deployment]
... recipe = ftw.recipe.deployment
... create-filebeat-config = false
... """)

And then run buildout again:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Uninstalling deployment.
Updating instance1.
Installing deployment.
<BLANKLINE>

We should NOT have a filebeat config for our deployment:

>>> import os
>>> os.path.isfile(os.path.join(sample_buildout, 'etc', 'filebeat.d', 'sample-buildout.yml'))
False

If we mistakenly add duplicate plone.recipe.zope2instance instances to parts:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance
... parts = instance1 instance1 instance1 instance1 instance1 deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running buildout will print a warning:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
ftw.recipe.deployment:
##########################################################
# Duplicate zope parts found!
# ['instance1', 'instance1', 'instance1', 'instance1', 'instance1']
# This could be caused by extending from multiple zeoclients/
# .cfg files instead of just one, and should be avoided.
# To prevent generating invalid logrotate configs, duplicate
# parts will be skipped, but you should still fix this issue.
##########################################################
<BLANKLINE>
Uninstalling deployment.
Updating instance1.
Installing deployment.
<BLANKLINE>

And we still end up with a valid logrotate config (no duplicated paths):

>>> cat(sample_buildout, 'etc', 'logrotate.d', 'sample-buildout')
... #doctest: -NORMALIZE_WHITESPACE
/sample-buildout/var/log/instance1.log
/sample-buildout/var/log/instance1-Z2.log {
    sharedscripts
    postrotate
        /bin/kill -SIGUSR2 `cat /sample-buildout/var/instance1.pid 2>/dev/null` >/dev/null 2>&1 || true
    endscript
}
/sample-buildout/var/log/instance1-json.log {
    copytruncate
    missingok
}

Let’s also add a zeo part. Thus we first need a fake plone.recipe.zeoserver recipe:

>>> mkdir(sample_buildout, 'plone.recipe.zeoserver')
>>> write(sample_buildout, 'plone.recipe.zeoserver', 'zeoserver.py',
... """
... import os, zc.buildout
...
... class Recipe(object):
...
...     def __init__(self, buildout, name, options):
...         self.name, self.options = name, options
...         options['zeo-log'] = os.path.join(
...                               buildout['buildout']['directory'],
...                               'var', 'log', self.name + '.log'
...                               )
...         self.storage_number = options.get('storage-number', '1')
...         self.blob_storage = options.get('blob-storage', '')
...
...     def install(self):
...         return tuple()
...
...     def update(self):
...         pass
... """)
>>> write(sample_buildout, 'plone.recipe.zeoserver', 'setup.py',
... """
... from setuptools import setup
...
... setup(
...     name = "plone.recipe.zeoserver",
...     entry_points = {'zc.buildout': ['default = zeoserver:Recipe']},
...     )
... """)
>>> write(sample_buildout, 'plone.recipe.zeoserver', 'README.txt', " ")

Create a buildout with multiple instance parts and a zeo part:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver
... parts = instance1 instance2 zeo deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [instance2]
... <= instance1
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Updating instance1.
Installing instance2.
Installing zeo.
Updating deployment.
<BLANKLINE>

Verify the contents of the logrotate configuration file:

>>> cat(sample_buildout, 'etc', 'logrotate.d', 'sample-buildout')
/sample-buildout/var/log/instance1.log
/sample-buildout/var/log/instance1-Z2.log {
    sharedscripts
    postrotate
        /bin/kill -SIGUSR2 `cat /sample-buildout/var/instance1.pid 2>/dev/null` >/dev/null 2>&1 || true
    endscript
}
/sample-buildout/var/log/instance2.log
/sample-buildout/var/log/instance2-Z2.log {
    sharedscripts
    postrotate
        /bin/kill -SIGUSR2 `cat /sample-buildout/var/instance2.pid 2>/dev/null` >/dev/null 2>&1 || true
    endscript
}
/sample-buildout/var/log/zeo.log {
    copytruncate
}
/sample-buildout/var/log/instance1-json.log {
    copytruncate
    missingok
}
/sample-buildout/var/log/instance2-json.log {
    copytruncate
    missingok
}

Verify the zeo run control script:

>>> cat(sample_buildout, 'bin', 'rc-zeo')
#!/bin/sh
<BLANKLINE>
# chkconfig: 345 85 15
# description: Starts ZEO server
<BLANKLINE>
START_SCRIPT="/sample-buildout/bin/zeo"
<BLANKLINE>
[ -f $START_SCRIPT ] || exit 1
<BLANKLINE>
# Source function library.
. /etc/init.d/functions
<BLANKLINE>
RETVAL=0
<BLANKLINE>
if [ $(whoami) != "root" ]; then
    echo "You must be root."
    exit 1
fi
<BLANKLINE>
case $1 in
    start|stop)
        su zope -c "$START_SCRIPT $*" </dev/null
        RETVAL=$?
        if [ $RETVAL -eq 0 ]
        then
            echo_success
        else
            echo_failure
        fi
        echo
        ;;
    restart)
        ${0} stop
        sleep 1
        ${0} start
        ;;
    *)
        su zope -c "$START_SCRIPT $*" </dev/null
        ;;
esac

Verify the run control script for instance 2:

>>> cat(sample_buildout, 'bin', 'rc-instance2')
#!/bin/sh
<BLANKLINE>
# chkconfig: 345 90 10
# description: Starts Zope
<BLANKLINE>
START_SCRIPT="/sample-buildout/bin/instance2"
<BLANKLINE>
[ -f $START_SCRIPT ] || exit 1
<BLANKLINE>
# Source function library.
. /etc/init.d/functions
<BLANKLINE>
RETVAL=0
<BLANKLINE>
if [ $(whoami) != "root" ]; then
    echo "You must be root."
    exit 1
fi
<BLANKLINE>
case $1 in
    start|stop)
        su zope -c "$START_SCRIPT $*" </dev/null
        RETVAL=$?
        if [ $RETVAL -eq 0 ]
        then
            echo_success
        else
            echo_failure
        fi
        echo
        ;;
    restart)
        ${0} stop
        sleep 1
        ${0} start
        ;;
    *)
        su zope -c "$START_SCRIPT $*" </dev/null
        ;;
esac

We should also have a packall script for packing all databases:

>>> cat(sample_buildout, 'bin', 'packall')
#!/bin/sh
/sample-buildout/bin/zeopack -S 1 -B /sample-buildout/var/blobstorage \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed Data (blobstorage)" >> /sample-buildout/var/log/pack.log

We can specify the user that should be used to run processes:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver
... parts = instance1 instance2 zeo deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [instance2]
... <= instance1
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [deployment]
... recipe = ftw.recipe.deployment
... rc-user = plone
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Uninstalling deployment.
Updating instance1.
Updating instance2.
Updating zeo.
Installing deployment.
<BLANKLINE>

Verify the run control script for instance 1:

>>> cat(sample_buildout, 'bin', 'rc-instance1')
#!/bin/sh
<BLANKLINE>
# chkconfig: 345 90 10
# description: Starts Zope
<BLANKLINE>
START_SCRIPT="/sample-buildout/bin/instance1"
<BLANKLINE>
[ -f $START_SCRIPT ] || exit 1
<BLANKLINE>
# Source function library.
. /etc/init.d/functions
<BLANKLINE>
RETVAL=0
<BLANKLINE>
if [ $(whoami) != "root" ]; then
    echo "You must be root."
    exit 1
fi
<BLANKLINE>
case $1 in
    start|stop)
        su plone -c "$START_SCRIPT $*" </dev/null
        RETVAL=$?
        if [ $RETVAL -eq 0 ]
        then
            echo_success
        else
            echo_failure
        fi
        echo
        ;;
    restart)
        ${0} stop
        sleep 1
        ${0} start
        ;;
    *)
        su plone -c "$START_SCRIPT $*" </dev/null
        ;;
esac

Before we can add a supervisor part we need a fake recipe for it:

>>> mkdir(sample_buildout, 'collective.recipe.supervisor')
>>> write(sample_buildout, 'collective.recipe.supervisor', 'supervisor.py',
... """
... import os, zc.buildout
...
... class Recipe(object):
...
...     def __init__(self, buildout, name, options):
...         pass
...
...     def install(self):
...         return tuple()
...
...     def update(self):
...         pass
... """)
>>> write(sample_buildout, 'collective.recipe.supervisor', 'setup.py',
... """
... from setuptools import setup
...
... setup(
...     name = "collective.recipe.supervisor",
...     entry_points = {'zc.buildout': ['default = supervisor:Recipe']},
...     )
... """)
>>> write(sample_buildout, 'collective.recipe.supervisor', 'README.txt',
... " ")

Create a buildout with a supervisor part:

>>> write('buildout.cfg',
... """
... [buildout]
... develop =
...     plone.recipe.zope2instance
...     plone.recipe.zeoserver
...     collective.recipe.supervisor
... parts = instance1 instance2 zeo supervisor deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [instance2]
... <= instance1
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [supervisor]
... recipe = collective.recipe.supervisor
...
... [deployment]
... recipe = ftw.recipe.deployment
... rc-user = plone
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.supervisor'
Updating instance1.
Updating instance2.
Updating zeo.
Installing supervisor.
Updating deployment.
<BLANKLINE>

Verify the supervisor control script:

>>> cat(sample_buildout, 'bin', 'rc-supervisor')
#!/bin/sh
<BLANKLINE>
# chkconfig: 345 90 10
# description: Starts supervisor
<BLANKLINE>
SUPERVISORCTL="/sample-buildout/bin/supervisorctl"
SUPERVISORD="/sample-buildout/bin/supervisord"
<BLANKLINE>
[ -f $SUPERVISORD ] || exit 1
[ -f $SUPERVISORCTL ] || exit 1
<BLANKLINE>
# Source function library.
. /etc/init.d/functions
<BLANKLINE>
RETVAL=0
<BLANKLINE>
if [ $(whoami) != "root" ]; then
    echo "You must be root."
    exit 1
fi
<BLANKLINE>
start() {
    echo -n "Starting supervisor: "
    su plone -c "$SUPERVISORD"
    RETVAL=$?
    if [ $RETVAL -eq 0 ]; then
        echo_success
    else
        echo_failure
    fi
    return $RETVAL
}
<BLANKLINE>
stop() {
    echo -n "Stopping supervisor: "
    su plone -c "$SUPERVISORCTL shutdown"
    RETVAL=$?
    if [ $RETVAL -eq 0 ]; then
        echo_success
    else
        echo_failure
    fi
    return $RETVAL
}
<BLANKLINE>
status() {
    su plone -c "$SUPERVISORCTL status"
}
<BLANKLINE>
case "$1" in
    start)
        start
        ;;
<BLANKLINE>
    stop)
        stop
        ;;
<BLANKLINE>
    restart)
        stop
        start
        ;;
<BLANKLINE>
    status)
        status
        ;;
esac
<BLANKLINE>
exit $REVAL

We can provide some additional logrotate options:

>>> write('buildout.cfg',
... """
... [buildout]
... develop =
...     plone.recipe.zope2instance
...     plone.recipe.zeoserver
...     collective.recipe.supervisor
...
... parts = instance1 deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [deployment]
... recipe = ftw.recipe.deployment
... logrotate-options =
...     rotate 4
...     weekly
...     missingok
...     notifempty
...     nomail
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.supervisor'
Uninstalling deployment.
Uninstalling supervisor.
Uninstalling zeo.
Uninstalling instance2.
Updating instance1.
Installing deployment.
<BLANKLINE>

Verify that the file contains our logrotate options:

>>> cat(sample_buildout, 'etc', 'logrotate.d', 'sample-buildout')
... #doctest: -NORMALIZE_WHITESPACE
/sample-buildout/var/log/instance1.log
/sample-buildout/var/log/instance1-Z2.log {
    rotate 4
    weekly
    missingok
    notifempty
    nomail
    sharedscripts
    postrotate
        /bin/kill -SIGUSR2 `cat /sample-buildout/var/instance1.pid 2>/dev/null` >/dev/null 2>&1 || true
    endscript
}
/sample-buildout/var/log/instance1-json.log {
    rotate 4
    weekly
    missingok
    notifempty
    nomail
    copytruncate
}

We can provide custom storage options:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver
... parts = instance1 zeo deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [zeo]
... recipe = plone.recipe.zeoserver
... storage-number = main
... blob-storage = blobstorage-main
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Uninstalling deployment.
Updating instance1.
Installing zeo.
Installing deployment.
<BLANKLINE>

Our packall script should contain the correct storage parameters:

>>> cat(sample_buildout, 'bin', 'packall')
#!/bin/sh
/sample-buildout/bin/zeopack -S main -B /sample-buildout/var/blobstorage-main \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed main (blobstorage-main)" >> /sample-buildout/var/log/pack.log

Let’s add a filestorage part. Thus we first need a fake collective.recipe.filestorage recipe:

>>> mkdir(sample_buildout, 'collective.recipe.filestorage')
>>> write(sample_buildout, 'collective.recipe.filestorage', 'filestorage.py',
... """
... import os, zc.buildout
...
... class Recipe(object):
...
...     def __init__(self, buildout, name, options):
...         self.name, self.options = name, options
...         self.subparts = options.get('parts', '').split()
...
...     def install(self):
...         return tuple()
...
...     def update(self):
...         pass
... """)
>>> write(sample_buildout, 'collective.recipe.filestorage', 'setup.py',
... """
... from setuptools import setup
...
... setup(
...     name = "collective.recipe.filestorage",
...     entry_points = {'zc.buildout': ['default = filestorage:Recipe']},
...     )
... """)
>>> write(sample_buildout, 'collective.recipe.filestorage', 'README.txt', " ")

Create a buildout with a filestorage part:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver collective.recipe.filestorage
... parts = instance1 zeo filestorage deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [filestorage]
... recipe = collective.recipe.filestorage
... parts = storage1
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.filestorage'
Uninstalling zeo.
Updating instance1.
Installing zeo.
Installing filestorage.
Updating deployment.
<BLANKLINE>

Our packall script should contain pack commands for all storages:

>>> cat(sample_buildout, 'bin', 'packall')
#!/bin/sh
/sample-buildout/bin/zeopack -S 1 -B /sample-buildout/var/blobstorage  \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed Data (blobstorage)" >> /sample-buildout/var/log/pack.log
/sample-buildout/bin/zeopack -S storage1 \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed storage1" >> /sample-buildout/var/log/pack.log

Let’s create a buildout with multiple filestorages and blobs:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver collective.recipe.filestorage
... parts = instance1 zeo filestorage deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [filestorage]
... recipe = collective.recipe.filestorage
... parts = storage1 storage2
... blob-storage = var/blobstorage-%(fs_part_name)s
... zeo-storage = %(fs_part_name)s_storage
...
... [deployment]
... recipe = ftw.recipe.deployment
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.filestorage'
Uninstalling filestorage.
Updating instance1.
Updating zeo.
Installing filestorage.
Updating deployment.
<BLANKLINE>

Our packall script should contain pack commands for all storages:

>>> cat(sample_buildout, 'bin', 'packall')
#!/bin/sh
/sample-buildout/bin/zeopack -S 1 -B /sample-buildout/var/blobstorage  \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed Data (blobstorage)" >> /sample-buildout/var/log/pack.log
/sample-buildout/bin/zeopack -S storage1_storage -B /sample-buildout/var/blobstorage-storage1 \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed storage1_storage (blobstorage-storage1)" >> /sample-buildout/var/log/pack.log
/sample-buildout/bin/zeopack -S storage2_storage -B /sample-buildout/var/blobstorage-storage2 \
    && echo `date +%Y-%m-%dT%H:%M:%S%z` "packed storage2_storage (blobstorage-storage2)" >> /sample-buildout/var/log/pack.log

Create a buildout with the packall-symlink-directory option:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver collective.recipe.filestorage
... parts = instance1 zeo deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [deployment]
... recipe = ftw.recipe.deployment
... packall-symlink-directory = etc/zodbpack.d
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.filestorage'
Uninstalling deployment.
Uninstalling filestorage.
Updating instance1.
Updating zeo.
Installing deployment.
<BLANKLINE>

We should now have a symlink in the given directory:

>>> ls(sample_buildout, 'etc', 'zodbpack.d')
l  sample-buildout

Create a buildout with startup/shutdown directory option:

>>> write('buildout.cfg',
... """
... [buildout]
... develop = plone.recipe.zope2instance plone.recipe.zeoserver collective.recipe.supervisor
... parts = instance1 zeo supervisor deployment
...
... [instance1]
... recipe = plone.recipe.zope2instance
...
... [zeo]
... recipe = plone.recipe.zeoserver
...
... [supervisor]
... recipe = collective.recipe.supervisor
...
... [deployment]
... recipe = ftw.recipe.deployment
... startup-directory = etc/startup.d
... shutdown-directory = etc/shutdown.d
... """)

Running the buildout gives us:

>>> print system(buildout)
Develop: '/sample-buildout/plone.recipe.zope2instance'
Develop: '/sample-buildout/plone.recipe.zeoserver'
Develop: '/sample-buildout/collective.recipe.supervisor'
Uninstalling deployment.
Updating instance1.
Updating zeo.
Installing supervisor.
Installing deployment.

Verify the startup script:

>>> cat(sample_buildout, 'etc', 'startup.d', 'sample-buildout')
#!/bin/sh
/sample-buildout/bin/supervisord

Verify the shutdown script:

>>> cat(sample_buildout, 'etc', 'shutdown.d', 'sample-buildout')
#!/bin/sh
/sample-buildout/bin/supervisorctl shutdown

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ftw.recipe.deployment-1.4.3.tar.gz (25.8 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page