configured properly will happily send them to lots of people who
probably didn't want those emails.
-o Don't use the debian dak.conf, apt.conf, cron.* etc. as starting
+o Don't use the debian dak.conf, cron.* etc. as starting
points for your own configuration files, they're highly Debian
specific. Start from scratch and refer to the security.debian.org
config files (-security) as they're a better example for a private
o To process queue/:
- * dak process-unchecked - processes queue/unchecked
- * dak process-accepted - move files from queue/accepted into the pool (and database)
- * dak process-new - allows ftp administrator to processes queue/new and queue/byhand
+ * dak process-upload - processes queue/unchecked
+ * dak process-new - allows ftp administrator to process queue/new and queue/byhand
+ * dak process-policy - processes policy queues (including new and byhand)
o To generate indices files:
* dak dominate - removes obsolete packages from suites
- * dak generate-filelist - generates file lists for apt-ftparchive
+ * dak generate-packages-sources2 - generate Packages, Sources
* dak generate-releases - generates Release
o To clean things up:
* dak ls - shows information about package(s)
* dak queue-report - shows information about package(s) in queue/
* dak override - can show you individual override entries
+ * dak graph - creates some pretty graphs of queue sizes over time
Generic and useful, but only for those with existing archives
-------------------------------------------------------------
-o dak poolize - migrates packages from legacy locations to the pool
-o dak init-archive - initializes a projectb database from an exisiting archive
+o dak init-archive - initializes a projectb database from an existing archive
Generic but not overly useful (in normal use)
---------------------------------------------
-o dak dot-dak-decode - dumps info in .dak files
o dak import-users-from-passwd - sync PostgreSQL users with system users
o dak cruft-report - check for obsolete or duplicated packages
o dak init-dirs - directory creation in the initial setup of an archive
o dak check-archive - various sanity checks of the database and archive
-o dak control-overrides - manpiulates/list override entries
+o dak control-overrides - manipulates/lists override entries
o dak control-suite - removes/adds/lists package(s) from/to/for a suite
o dak stats - produces various statistics
o dak find-null-maintainers - checks for users with no packages in the archive
----------------------------
o dak security-install - wrapper for Debian security team
-o dak clean-proposed-updates - removes obsolete .changes files from proposed-updates
-o dak check-proposed-updates - basic dependency checking for proposed-updates
-o dak reject-proposed-updates - manually reject packages from proposed-updates
o dak import-ldap-fingerprints - syncs fingerprint and uid information with a debian.org LDAP DB
Very Incomplete or otherwise not generally useful
-------------------------------------------------
o dak init-db - currently only initializes a DB from a dak.conf config file
-o dak compare-suites - looks for version descrepancies that shouldn't exist in many
- archives
o dak check-overrides - override cruft checker that doesn't work well with New Incoming
Scripts invoked by other scripts
--------------------------------
o dak examine-package - invoked by 'dak process-new' to "check" NEW packages
-o dak symlink-dists - invoked by 'dak poolize' to determine packages still in legacy locations
How do I get started?
=====================
-[Very incomplete - FIXME]
-
-o Write your own dak.conf and apt.conf files. dak looks for those
- config files in /etc/dak/. /etc/dak/dak.conf can define
- alternative configuration files with Config::host::DakConfig and
- Config::host::AptConfig (where "host" is the fully qualified domain
- name of your machine).
-o Create a PostgreSQL database on the host given in dak.conf's DB::Host
- with the name specified in DB::Name.
-o Run 'dak init-dirs': this will create all directories which are specified in
- dak.conf and apt.conf.
-o If you have an existing archive:
- * Run 'dak init-archive'
- otherwise:
- * Create the table structure. init_pool.sql contains all SQL statements
- which are needed for this. After changing all occurences of "projectb"
- to the name of your database (as defined in DB::Name) you can run:
- psql <DB::Name> < init_pool.sql
- * Create the following groups in postgres: ftpmaster ftpteam ftptrainee
- * Run 'dak update-db' to upgrade the database schema.
- * Run 'dak init-db': it will populate your database with the values from
- dak.conf and apt.conf.
- * Run 'psql <DB::Name> < add_constraints.sql'.
-o Copy all templates from the "templates" directory to to the directory
- specified in Dir::Templates, and adapt them to your distribution.
-
+Please refer to setup/README for instructions how to setup dak.
projects / dak.git / blobdiff