5 - Implement "group" filtering using blanche for limited rollouts.
7 - Remove "already migrated" cruft that will accumulate if we do small
8 --limit and then increase.
9 - Make single user mass-migrate work when not logged in as root
11 - Make parallel-find.pl use `sudo -u username git describe --tags`
12 to determine version. Make parallel-find.pl have this have greater
13 precedence. This also means, however, that we get
14 full mediawiki-1.2.3-2-abcdef names (Have patch, pending testing and commit)
15 - Make deployed installer use 'wizard install' /or/ do a migration
16 after doing a normal install (the latter makes it easier
19 - Better error message if daemon/scripts-security-upd
20 is not on scripts-security-upd list
22 - Custom merge algo: absolute php.ini symlinks to relative symlinks
23 - Custom merge algo: re-constitute AdminSettings.php if missing. It looks
24 like this is the case for most 1.5.8 installs (check what the merges
25 do in both directions). All 1.11.0 installs except four have
26 the other (check diff -u with all in /root)
28 - Redo Wordpress conversion, with an eye for automating everything
29 possible (such as downloading the tarball and unpacking)
31 - Genericize callAsUser and drop_priviledges in shell
32 - Summary script should be more machine friendly, and should not
33 output summary charts when I increase specificity
35 Some other stuff to do in your copious free time:
36 - Summary script should do something intelligent when distinguishing
37 between old-style and new-style installs
38 - Check how many autoinstalls are missing w bits for
39 daemon.scripts (this would need pyafs)
40 - Make scripts AFS patch advertise its existence so we can check for it.
41 (This might be otherwise possible using `fs sysname`)
42 - Make 'wizard summary' generate nice pretty graphs of installs by date
43 (more histograms, will need to check actual .scripts-version files.)
44 - It should be able to handle installs like Django where there's a component
45 that gets installed in web_scripts and another directory that gets installed
48 PULLING OUT CONFIGURATION FILES IN AN AUTOMATED MANNER
50 advancedpoll: Template file to fill out
51 django: Noodles of template files
52 gallery2: Multistage install process
54 mediawiki: One-step install process
55 phpbb: Multistage install process
56 phpical: Template file
59 wordpress: Multistage install process
61 PHILOSOPHY ABOUT LOGGING
63 Logging is most useful when performing a mass run. This
64 includes things such as mass-migration as well as when running
65 summary reports. An interesting property about mass-migration
66 or mass-upgrade, however, is that if they fail, they are
67 idempotent, so an individual case can be debugged simply running
68 the single-install equivalent with --debug on. (This, indeed,
69 may be easier to do than sifting through a logfile).
71 It is a different story when you are running a summary report:
72 you are primarily bound by your AFS cache and how quickly you can
73 iterate through all of the autoinstalls. Checking if a file
74 exists on a cold AFS cache may
75 take several minutes to perform; on a hot cache the same report
76 may take a mere 3 seconds. When you get to more computationally
77 expensive calculations, however, even having a hot AFS cache
78 is not enough to cut down your runtime.
80 There are certain calculations that someone may want to be
81 able to perform on manipulated data. As such, this data should
82 be cached on disk, if the process for extracting this data takes
83 a long time. Also, for usability sake, Wizard should generate
84 the common case reports.
86 Ensuring that machine parseable reports are made, and then making
87 the machinery to reframe this data, increases complexity. Therefore,
88 the recommendation is to assume that if you need to run iteratively,
89 you'll have a hot AFS cache at your fingerprints, and if that's not
90 fast enough, then cache the data.
92 COMMIT MESSAGE FIELDS:
94 Installed-by: username@hostname
95 Pre-commit-by: Real Name <username@mit.edu>
96 Upgraded-by: Real Name <username@mit.edu>
97 Migrated-by: Real Name <username@mit.edu>
98 Wizard-revision: abcdef1234567890
99 Wizard-args: /wizard/bin/wizard foo bar baz
103 Committer: Real Name <username@mit.edu>
104 Author: lockername locker <lockername@scripts.mit.edu>
108 - It is not expected or required for update scripts to exist for all
109 intervening versions that were present pre-migration; only for it
110 to work on the most recent migration.
112 - Currently all repositories are initialized with --shared, which
113 means they have basically ~no space footprint. However, it
114 also means that /mit/scripts/wizard/srv MUST NOT lose revs after
117 - Full fledged logging options. Namely:
118 x all loggers (delay implementing this until we actually have debug stmts)
120 - debug => loglevel = DEBUG
122 - default is WARNING (see below for exception)
123 - verbose => loglevel = INFO
124 x file logger (creates a dir and lots of little logfiles)
126 - log-file => loglevel = INFO
130 * Some parts of the infrastructure will not be touched, although I plan
131 on documenting them. Specifically, we will be keeping:
133 - parallel-find.pl, and the resulting
134 /mit/scripts/.htaccess/scripts/sec-tools/store/scriptslist
136 * The new procedure for generating an update is as follows:
137 (check out the mass-migration instructions for something in this spirit,
138 although uglier in some ways; A indicates the step /should/ be automated)
140 0. ssh into not-backward, temporarily give the daemon.scripts-security-upd
141 bits by blanching it on system:scripts-security-upd, and run parallel-find.pl
143 1. Have the Git repository and working copy for the project on hand.
145 /- wizard prepare-pristine --
147 A 2. Checkout the pristine branch
149 A 3. Remove all files from the working copy. Use `wipe-working-dir`
151 A 4. Download the new tarball
153 A 5. Extract the tarball over the working copy (`cp -R a/. b` works well,
154 remember that the working copy is empty; this needs some intelligent
157 A 6. Check for empty directories and add stub files as necessary.
158 Use `preserve-empty-dir`
162 7. Git add it all, and then commit as a new pristine version (v1.2.3)
164 8. Checkout the master branch
166 9. [FOR EXISTING REPOSITORIES]
167 Merge the pristine branch in. Resolve any conflicts that our
168 patches have with new changes. Do NOT let Git auto-commit it
169 with --no-commit (otherwise, you want to git commit --amend
170 to keep our history clean
172 [FOR NEW REPOSITORIES]
173 Check if any patches are needed to make the application work
174 on Scripts (ideally, it shouldn't.
176 /- wizard prepare-new --
178 Currently not used for anything besides parallel-find.pl, but
179 we reserve the right to place files in here in the future.
182 A echo "Deny from all" > .scripts/.htaccess
186 10. Check if there are any special update procedures, and update
187 the wizard.app.APPNAME module accordingly (or create it, if
190 11. Run 'wizard prepare-config' on a scripts server while in a checkout
191 of this newest version. This will prepare a new version of the
192 configuration file based on the application's latest installer.
193 Manually merge back in any custom changes we may have made.
194 Check if any of the regular expressions need tweaking by inspecting
195 the configuration files for user-specific gunk, and modify
196 wizard.app.APPNAME accordingly.
198 12. Commit your changes, and tag as v1.2.3-scripts (or scripts2, if
199 you are amending an install without an upstream changes)
201 NOTE: These steps should be run on a scripts server
203 13. Test the new update procedure using our test scripts. See integration
204 tests for more information on how to do this.
206 http://scripts.mit.edu/wizard/testing.html#acceptance-tests
208 GET APPROVAL BEFORE PROCEEDING ANY FURTHER
210 NOTE: The following commands are to be run on not-backward.mit.edu.
211 You'll need to add daemon.scripts-security-upd to
212 scripts-security-upd to get bits to do this. Make sure you remove
213 these bits when you're done.
215 A 14. Run `wizard research appname`
216 which uses Git commands to check how many
217 working copies apply the change cleanly, and writes out a logfile
218 with the working copies that don't apply cleanly. It also tells
219 us about "corrupt" working copies, i.e. working copies that
220 have over a certain threshold of changes.
222 A 15. Run `wizard mass-upgrade appname`, which applies the update to all working
223 copies possible, and sends mail to users to whom the working copy
224 did not apply cleanly.
226 16. Run parallel-find.pl to update our inventory
228 * For mass importing into the repository, there are a few extra things:
230 * Many applications had patches associated with them. Be sure to
231 apply them, so later merges work better.
233 # the following operation might require -p1
234 patch -p0 < ../app-1.2.3/app-1.2.3.patch # [FIDDLY BIT]
236 * When running updates, if the patch has changed you will have to
237 do a special procedure for your merge:
239 git checkout pristine
240 # NOTE: Now, the tricky part (this is different from a real update)
241 git symbolic-ref HEAD refs/heads/master
242 # NOTE: Now, we think we're on the master branch, but we have
243 # pristine copy checked out
244 # NOTE: -p0 might need to be twiddled
245 patch -p0 < ../app-1.2.3/app-1.2.3.patch
247 # reconstitute .scripts directory
248 git checkout v1.2.2-scripts -- .scripts
250 # NOTE: Fake the merge
251 git rev-parse pristine > .git/MERGE_HEAD
253 You could also just try your luck with a manual merge using the patch
256 * The repository for a given application will contain the following files:
258 - The actual application's files, as from the official tarball
260 - A .scripts directory, with the intent of holding Scripts specific files
261 if they become necessary.
263 * .scripts/lock (generated) which locks an autoinstall during upgrade