blog/tags/bootsystem/index.html

   1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
   2           "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
   3 <html xmlns="http://www.w3.org/1999/xhtml" dir="ltr">
   4   <head>
   5     <meta http-equiv="Content-Type" content="text/html;charset=utf-8" />
   6     <title>Petter Reinholdtsen: Entries Tagged bootsystem</title>
   7     <link rel="stylesheet" type="text/css" media="screen" href="http://people.skolelinux.org/pere/blog/style.css" />
   8     <link rel="stylesheet" type="text/css" media="screen" href="http://people.skolelinux.org/pere/blog/vim.css" />
   9     <link rel="alternate" title="RSS Feed" href="bootsystem.rss" type="application/rss+xml" />
  10   </head>
  11   <body>
  12     <div class="title">
  13  <h1>
  14      <a href="http://people.skolelinux.org/pere/blog/">Petter Reinholdtsen</a>
  15
  16  </h1>
  17
  18 </div>
  19
  20
  21     <h3>Entries tagged "bootsystem".</h3>
  22
  23     <div class="entry">
  24       <div class="title">
  25         <a href="http://people.skolelinux.org/pere/blog/Debian_boots_quicker_and_quicker.html">Debian boots quicker and quicker</a>
  26       </div>
  27       <div class="date">
  28         24th June 2009
  29       </div>
  30       <div class="body">
  31         <p>I spent Monday and tuesday this week in London with a lot of the
  32 people involved in the boot system on Debian and Ubuntu, to see if we
  33 could find more ways to speed up the boot system.  This was an Ubuntu
  34 funded
  35 <a href="https://wiki.ubuntu.com/FoundationsTeam/BootPerformance/DebianUbuntuSprint">developer
  36 gathering</a>. It was quite productive.  We also discussed the future
  37 of boot systems, and ways to handle the increasing number of boot
  38 issues introduced by the Linux kernel becoming more and more
  39 asynchronous and event base.  The Ubuntu approach using udev and
  40 upstart might be a good way forward.  Time will show.</p>
  41
  42 <p>Anyway, there are a few ways at the moment to speed up the boot
  43 process in Debian.  All of these should be applied to get a quick
  44 boot:</p>
  45
  46 <ul>
  47
  48 <li>Use dash as /bin/sh.</li>
  49
  50 <li>Disable the init.d/hwclock*.sh scripts and make sure the hardware
  51    clock is in UTC.</li>
  52
  53 <li>Install and activate the insserv package to enable
  54    <a href="http://wiki.debian.org/LSBInitScripts/DependencyBasedBoot">dependency
  55    based boot sequencing</a>, and enable concurrent booting.</li>
  56
  57 </ul>
  58
  59 These points are based on the Google summer of code work done by
  60 <a href="http://initscripts-ng.alioth.debian.org/soc2006-bootsystem/">Carlos
  61 Villegas</a>.
  62
  63 <p>Support for makefile-style concurrency during boot was uploaded to
  64 unstable yesterday.  When we tested it, we were able to cut 6 seconds
  65 from the boot sequence.  It depend on very correct dependency
  66 declaration in all init.d scripts, so I expect us to find edge cases
  67 where the dependences in some scripts are slightly wrong when we start
  68 using this.</p>
  69
  70 <p>On our IRC channel for this effort, #pkg-sysvinit, a new idea was
  71 introduced by Raphael Geissert today, one that could affect the
  72 startup speed as well.  Instead of starting some scripts concurrently
  73 from rcS.d/ and another set of scripts from rc2.d/, it would be
  74 possible to run a of them in the same process.  A quick way to test
  75 this would be to enable insserv and run 'mv /etc/rc2.d/S* /etc/rcS.d/;
  76 insserv'.  Will need to test if that work. :)</p>
  77
  78       </div>
  79       <div class="tags">
  80
  81
  82         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
  83
  84
  85       </div>
  86     </div>
  87     <div class="padding"></div>
  88
  89     <div class="entry">
  90       <div class="title">
  91         <a href="http://people.skolelinux.org/pere/blog/Taking_over_sysvinit_development.html">Taking over sysvinit development</a>
  92       </div>
  93       <div class="date">
  94         22nd July 2009
  95       </div>
  96       <div class="body">
  97         <p>After several years of frustration with the lack of activity from
  98 the existing sysvinit upstream developer, I decided a few weeks ago to
  99 take over the package and become the new upstream.  The number of
 100 patches to track for the Debian package was becoming a burden, and the
 101 lack of synchronization between the distribution made it hard to keep
 102 the package up to date.</p>
 103
 104 <p>On the new sysvinit team is the SuSe maintainer Dr. Werner Fink,
 105 and my Debian co-maintainer Kel Modderman.  About 10 days ago, I made
 106 a new upstream tarball with version number 2.87dsf (for Debian, SuSe
 107 and Fedora), based on the patches currently in use in these
 108 distributions.  We Debian maintainers plan to move to this tarball as
 109 the new upstream as soon as we find time to do the merge.  Since the
 110 new tarball was created, we agreed with Werner at SuSe to make a new
 111 upstream project at <a href="http://savannah.nongnu.org/">Savannah</a>, and continue
 112 development there.  The project is registered and currently waiting
 113 for approval by the Savannah administrators, and as soon as it is
 114 approved, we will import the old versions from svn and continue
 115 working on the future release.</p>
 116
 117 <p>It is a bit ironic that this is done now, when some of the involved
 118 distributions are moving to upstart as a syvinit replacement.</p>
 119
 120       </div>
 121       <div class="tags">
 122
 123
 124         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>, <a href="http://people.skolelinux.org/pere/blog/tags/nuug">nuug</a>.
 125
 126
 127       </div>
 128     </div>
 129     <div class="padding"></div>
 130
 131     <div class="entry">
 132       <div class="title">
 133         <a href="http://people.skolelinux.org/pere/blog/Debian_has_switched_to_dependency_based_boot_sequencing.html">Debian has switched to dependency based boot sequencing</a>
 134       </div>
 135       <div class="date">
 136         27th July 2009
 137       </div>
 138       <div class="body">
 139         <p>Since this evening, with the upload of sysvinit version 2.87dsf-2,
 140 and the upload of insserv version 1.12.0-10 yesterday, Debian unstable
 141 have been migrated to using dependency based boot sequencing.  This
 142 conclude work me and others have been doing for the last three days.
 143 It feels great to see this finally part of the default Debian
 144 installation.  Now we just need to weed out the last few problems that
 145 are bound to show up, to get everything ready for Squeeze.</p>
 146
 147 <p>The next step is migrating /sbin/init from sysvinit to upstart, and
 148 fixing the more fundamental problem of handing the event based
 149 non-predictable kernel in the early boot.</p>
 150
 151       </div>
 152       <div class="tags">
 153
 154
 155         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>, <a href="http://people.skolelinux.org/pere/blog/tags/nuug">nuug</a>.
 156
 157
 158       </div>
 159     </div>
 160     <div class="padding"></div>
 161
 162     <div class="entry">
 163       <div class="title">
 164         <a href="http://people.skolelinux.org/pere/blog/Parallellizing_the_boot_in_Debian_Squeeze___ready_for_wider_testing.html">Parallellizing the boot in Debian Squeeze - ready for wider testing</a>
 165       </div>
 166       <div class="date">
 167          6th May 2010
 168       </div>
 169       <div class="body">
 170         <p>These days, the init.d script dependencies in Squeeze are quite
 171 complete, so complete that it is actually possible to run all the
 172 init.d scripts in parallell based on these dependencies.  If you want
 173 to test your Squeeze system, make sure
 174 <a href="http://wiki.debian.org/LSBInitScripts/DependencyBasedBoot">dependency
 175 based boot sequencing</a> is enabled, and add this line to
 176 /etc/default/rcS:</p>
 177
 178 <blockquote><pre>
 179 CONCURRENCY=makefile
 180 </pre></blockquote>
 181
 182 <p>That is it.  It will cause sysv-rc to use the startpar tool to run
 183 scripts in parallel using the dependency information stored in
 184 /etc/init.d/.depend.boot, /etc/init.d/.depend.start and
 185 /etc/init.d/.depend.stop to order the scripts.  Startpar is configured
 186 to try to start the kdm and gdm scripts as early as possible, and will
 187 start the facilities required by kdm or gdm as early as possible to
 188 make this happen.</p>
 189
 190 <p>Give it a try, and see if you like the result.  If some services
 191 fail to start properly, it is most likely because they have incomplete
 192 init.d script dependencies in their startup script (or some of their
 193 dependent scripts have incomplete dependencies).  Report bugs and get
 194 the package maintainers to fix it. :)</p>
 195
 196 <p>Running scripts in parallel could be the default in Debian when we
 197 manage to get the init.d script dependencies complete and correct.  I
 198 expect we will get there in Squeeze+1, if we get manage to test and
 199 fix the remaining issues.</p>
 200
 201 <p>If you report any problems with dependencies in init.d scripts to
 202 the BTS, please usertag the report to get it to show up at
 203 <a href="http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=initscripts-ng-devel@lists.alioth.debian.org">the
 204 list of usertagged bugs related to this</a>.</p>
 205
 206       </div>
 207       <div class="tags">
 208
 209
 210         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 211
 212
 213       </div>
 214     </div>
 215     <div class="padding"></div>
 216
 217     <div class="entry">
 218       <div class="title">
 219         <a href="http://people.skolelinux.org/pere/blog/systemd__an_interesting_alternative_to_upstart.html">systemd, an interesting alternative to upstart</a>
 220       </div>
 221       <div class="date">
 222         13th May 2010
 223       </div>
 224       <div class="body">
 225         <p>The last few days a new boot system called
 226 <a href="http://www.freedesktop.org/wiki/Software/systemd">systemd</a>
 227 has been
 228 <a href="http://0pointer.de/blog/projects/systemd.html">introduced</a>
 229
 230 to the free software world.  I have not yet had time to play around
 231 with it, but it seem to be a very interesting alternative to
 232 <a href="http://upstart.ubuntu.com/">upstart</a>, and might prove to be
 233 a good alternative for Debian when we are able to switch to an event
 234 based boot system.  Tollef is
 235 <a href="http://bugs.debian.org/580814">in the process</a> of getting
 236 systemd into Debian, and I look forward to seeing how well it work.  I
 237 like the fact that systemd handles init.d scripts with dependency
 238 information natively, allowing them to run in parallel where upstart
 239 at the moment do not.</p>
 240
 241 <p>Unfortunately do systemd have the same problem as upstart regarding
 242 platform support.  It only work on recent Linux kernels, and also need
 243 some new kernel features enabled to function properly.  This means
 244 kFreeBSD and Hurd ports of Debian will need a port or a different boot
 245 system.  Not sure how that will be handled if systemd proves to be the
 246 way forward.</p>
 247
 248 <p>In the mean time, based on the
 249 <a href="http://lists.debian.org/debian-devel/2010/05/msg00122.html">input
 250 on debian-devel@</a> regarding parallel booting in Debian, I have
 251 decided to enable full parallel booting as the default in Debian as
 252 soon as possible (probably this weekend or early next week), to see if
 253 there are any remaining serious bugs in the init.d dependencies.  A
 254 new version of the sysvinit package implementing this change is
 255 already in experimental.  If all go well, Squeeze will be released
 256 with parallel booting enabled by default.</p>
 257
 258       </div>
 259       <div class="tags">
 260
 261
 262         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>, <a href="http://people.skolelinux.org/pere/blog/tags/nuug">nuug</a>.
 263
 264
 265       </div>
 266     </div>
 267     <div class="padding"></div>
 268
 269     <div class="entry">
 270       <div class="title">
 271         <a href="http://people.skolelinux.org/pere/blog/Parallellized_boot_is_now_the_default_in_Debian_unstable.html">Parallellized boot is now the default in Debian/unstable</a>
 272       </div>
 273       <div class="date">
 274         14th May 2010
 275       </div>
 276       <div class="body">
 277         <p>Since this evening, parallel booting is the default in
 278 Debian/unstable for machines using dependency based boot sequencing.
 279 Apparently the testing of concurrent booting has been wider than
 280 expected, if I am to believe the
 281 <a href="http://lists.debian.org/debian-devel/2010/05/msg00122.html">input
 282 on debian-devel@</a>, and I concluded a few days ago to move forward
 283 with the feature this weekend, to give us some time to detect any
 284 remaining problems before Squeeze is frozen.  If serious problems are
 285 detected, it is simple to change the default back to sequential boot.
 286 The upload of the new sysvinit package also activate a new upstream
 287 version.</p>
 288
 289 More information about
 290 <a href="http://wiki.debian.org/LSBInitScripts/DependencyBasedBoot">dependency
 291 based boot sequencing</a> is available from the Debian wiki.  It is
 292 currently possible to disable parallel booting when one run into
 293 problems caused by it, by adding this line to /etc/default/rcS:</p>
 294
 295 <blockquote><pre>
 296 CONCURRENCY=none
 297 </pre></blockquote>
 298
 299 <p>If you report any problems with dependencies in init.d scripts to
 300 the BTS, please usertag the report to get it to show up at
 301 <a href="http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=initscripts-ng-devel@lists.alioth.debian.org">the
 302 list of usertagged bugs related to this</a>.</p>
 303
 304       </div>
 305       <div class="tags">
 306
 307
 308         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian edu">debian edu</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 309
 310
 311       </div>
 312     </div>
 313     <div class="padding"></div>
 314
 315     <div class="entry">
 316       <div class="title">
 317         <a href="http://people.skolelinux.org/pere/blog/Parallellized_boot_seem_to_hold_up_well_in_Debian_testing.html">Parallellized boot seem to hold up well in Debian/testing</a>
 318       </div>
 319       <div class="date">
 320         27th May 2010
 321       </div>
 322       <div class="body">
 323         <p>A few days ago, parallel booting was enabled in Debian/testing.
 324 The feature seem to hold up pretty well, but three fairly serious
 325 issues are known and should be solved:
 326
 327 <p><ul>
 328
 329 <li>The wicd package seen to
 330 <a href="http://bugs.debian.org/508289">break NFS mounting</a> and
 331 <a href="http://bugs.debian.org/581586">network setup</a> when
 332 parallel booting is enabled.  No idea why, but the wicd maintainer
 333 seem to be on the case.</li>
 334
 335 <li>The nvidia X driver seem to
 336 <a href="http://bugs.debian.org/583312">have a race condition</a>
 337 triggered more easily when parallel booting is in effect.  The
 338 maintainer is on the case.</li>
 339
 340 <li>The sysv-rc package fail to properly enable dependency based boot
 341 sequencing (the shutdown is broken) when old file-rc users
 342 <a href="http://bugs.debian.org/575080">try to switch back</a> to
 343 sysv-rc.  One way to solve it would be for file-rc to create
 344 /etc/init.d/.legacy-bootordering, and another is to try to make
 345 sysv-rc more robust.  Will investigate some more and probably upload a
 346 workaround in sysv-rc to help those trying to move from file-rc to
 347 sysv-rc get a working shutdown.</li>
 348
 349 </ul></p>
 350
 351 <p>All in all not many surprising issues, and all of them seem
 352 solvable before Squeeze is released.  In addition to these there are
 353 some packages with bugs in their dependencies and run level settings,
 354 which I expect will be fixed in a reasonable time span.</p>
 355
 356 <p>If you report any problems with dependencies in init.d scripts to
 357 the BTS, please usertag the report to get it to show up at
 358 <a href="http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=initscripts-ng-devel@lists.alioth.debian.org">the
 359 list of usertagged bugs related to this</a>.</p>
 360
 361 <p>Update: Correct bug number to file-rc issue.</p>
 362
 363       </div>
 364       <div class="tags">
 365
 366
 367         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian edu">debian edu</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 368
 369
 370       </div>
 371     </div>
 372     <div class="padding"></div>
 373
 374     <div class="entry">
 375       <div class="title">
 376         <a href="http://people.skolelinux.org/pere/blog/KDM_fail_at_boot_with_NVidia_cards___and_no_one_try_to_fix_it_.html">KDM fail at boot with NVidia cards - and no one try to fix it?</a>
 377       </div>
 378       <div class="date">
 379          1st June 2010
 380       </div>
 381       <div class="body">
 382         <p>It is strange to watch how a bug in Debian causing KDM to fail to
 383 start at boot when an NVidia video card is used is handled.  The
 384 problem seem to be that the nvidia X.org driver uses a long time to
 385 initialize, and this duration is longer than kdm is configured to
 386 wait.</p>
 387
 388 <p>I came across two bugs related to this issue,
 389 <a href="http://bugs.debian.org/583312">#583312</a> initially filed
 390 against initscripts and passed on to nvidia-glx when it became obvious
 391 that the nvidia drivers were involved, and
 392 <a href="http://bugs.debian.org/524751">#524751</a> initially filed against
 393 kdm and passed on to src:nvidia-graphics-drivers for unknown reasons.</p>
 394
 395 <p>To me, it seem that no-one is interested in actually solving the
 396 problem nvidia video card owners experience and make sure the Debian
 397 distribution work out of the box for these users.  The nvidia driver
 398 maintainers expect kdm to be set up to wait longer, while kdm expect
 399 the nvidia driver maintainers to fix the driver to start faster, and
 400 while they wait for each other I guess the users end up switching to a
 401 distribution that work for them.  I have no idea what the solution is,
 402 but I am pretty sure that waiting for each other is not it.</p>
 403
 404 <p>I wonder why we end up handling bugs this way.</p>
 405
 406       </div>
 407       <div class="tags">
 408
 409
 410         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian edu">debian edu</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 411
 412
 413       </div>
 414     </div>
 415     <div class="padding"></div>
 416
 417     <div class="entry">
 418       <div class="title">
 419         <a href="http://people.skolelinux.org/pere/blog/Upstart_or_sysvinit___as_init_d_scripts_see_it.html">Upstart or sysvinit - as init.d scripts see it</a>
 420       </div>
 421       <div class="date">
 422          6th June 2010
 423       </div>
 424       <div class="body">
 425         <p>If Debian is to migrate to upstart on Linux, I expect some init.d
 426 scripts to migrate (some of) their operations to upstart job while
 427 keeping the init.d for hurd and kfreebsd.  The packages with such
 428 needs will need a way to get their init.d scripts to behave
 429 differently when used with sysvinit and with upstart.  Because of
 430 this, I had a look at the environment variables set when a init.d
 431 script is running under upstart, and when it is not.</p>
 432
 433 <p>With upstart, I notice these environment variables are set when a
 434 script is started from rcS.d/ (ignoring some irrelevant ones like
 435 COLUMNS):</p>
 436
 437 <blockquote><pre>
 438 DEFAULT_RUNLEVEL=2
 439 previous=N
 440 PREVLEVEL=
 441 RUNLEVEL=
 442 runlevel=S
 443 UPSTART_EVENTS=startup
 444 UPSTART_INSTANCE=
 445 UPSTART_JOB=rc-sysinit
 446 </pre></blockquote>
 447
 448 <p>With sysvinit, these environment variables are set for the same
 449 script.</p>
 450
 451 <blockquote><pre>
 452 INIT_VERSION=sysvinit-2.88
 453 previous=N
 454 PREVLEVEL=N
 455 RUNLEVEL=S
 456 runlevel=S
 457 </pre></blockquote>
 458
 459 <p>The RUNLEVEL and PREVLEVEL environment variables passed on from
 460 sysvinit are not set by upstart.  Not sure if it is intentional or not
 461 to not be compatible with sysvinit in this regard.</p>
 462
 463 <p>For scripts needing to behave differently when upstart is used,
 464 looking for the UPSTART_JOB environment variable seem to be a good
 465 choice.</p>
 466
 467       </div>
 468       <div class="tags">
 469
 470
 471         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 472
 473
 474       </div>
 475     </div>
 476     <div class="padding"></div>
 477
 478     <div class="entry">
 479       <div class="title">
 480         <a href="http://people.skolelinux.org/pere/blog/Automatic_upgrade_testing_from_Lenny_to_Squeeze.html">Automatic upgrade testing from Lenny to Squeeze</a>
 481       </div>
 482       <div class="date">
 483         11th June 2010
 484       </div>
 485       <div class="body">
 486         <p>The last few days I have done some upgrade testing in Debian, to
 487 see if the upgrade from Lenny to Squeeze will go smoothly.  A few bugs
 488 have been discovered and reported in the process
 489 (<a href="http://bugs.debian.org/585410">#585410</a> in nagios3-cgi,
 490 <a href="http://bugs.debian.org/584879">#584879</a> already fixed in
 491 enscript and <a href="http://bugs.debian.org/584861">#584861</a> in
 492 kdebase-workspace-data), and to get a more regular testing going on, I
 493 am working on a script to automate the test.</p>
 494
 495 <p>The idea is to create a Lenny chroot and use tasksel to install a
 496 Gnome or KDE desktop installation inside the chroot before upgrading
 497 it.  To ensure no services are started in the chroot, a policy-rc.d
 498 script is inserted.  To make sure tasksel believe it is to install a
 499 desktop on a laptop, the tasksel tests are replaced in the chroot
 500 (only acceptable because this is a throw-away chroot).</p>
 501
 502 <p>A naive upgrade from Lenny to Squeeze using aptitude dist-upgrade
 503 currently always fail because udev refuses to upgrade with the kernel
 504 in Lenny, so to avoid that problem the file /etc/udev/kernel-upgrade
 505 is created.  The bug report
 506 <a href="http://bugs.debian.org/566000">#566000</a> make me suspect
 507 this problem do not trigger in a chroot, but I touch the file anyway
 508 to make sure the upgrade go well.  Testing on virtual and real
 509 hardware have failed me because of udev so far, and creating this file
 510 do the trick in such settings anyway.  This is a
 511 <a href="http://www.linuxquestions.org/questions/debian-26/failed-dist-upgrade-due-to-udev-config_sysfs_deprecated-nonsense-804130/">known
 512 issue</a> and the current udev behaviour is intended by the udev
 513 maintainer because he lack the resources to rewrite udev to keep
 514 working with old kernels or something like that.  I really wish the
 515 udev upstream would keep udev backwards compatible, to avoid such
 516 upgrade problem, but given that they fail to do so, I guess
 517 documenting the way out of this mess is the best option we got for
 518 Debian Squeeze.</p>
 519
 520 <p>Anyway, back to the task at hand, testing upgrades.  This test
 521 script, which I call <tt>upgrade-test</tt> for now, is doing the
 522 trick:</p>
 523
 524 <blockquote><pre>
 525 #!/bin/sh
 526 set -ex
 527
 528 if [ "$1" ] ; then
 529     desktop=$1
 530 else
 531     desktop=gnome
 532 fi
 533
 534 from=lenny
 535 to=squeeze
 536
 537 exec &lt; /dev/null
 538 unset LANG
 539 mirror=http://ftp.skolelinux.org/debian
 540 tmpdir=chroot-$from-upgrade-$to-$desktop
 541 fuser -mv .
 542 debootstrap $from $tmpdir $mirror
 543 chroot $tmpdir aptitude update
 544 cat > $tmpdir/usr/sbin/policy-rc.d &lt;&lt;EOF
 545 #!/bin/sh
 546 exit 101
 547 EOF
 548 chmod a+rx $tmpdir/usr/sbin/policy-rc.d
 549 exit_cleanup() {
 550     umount $tmpdir/proc
 551 }
 552 mount -t proc proc $tmpdir/proc
 553 # Make sure proc is unmounted also on failure
 554 trap exit_cleanup EXIT INT
 555
 556 chroot $tmpdir aptitude -y install debconf-utils
 557
 558 # Make sure tasksel autoselection trigger.  It need the test scripts
 559 # to return the correct answers.
 560 echo tasksel tasksel/desktop multiselect $desktop | \
 561     chroot $tmpdir debconf-set-selections
 562
 563 # Include the desktop and laptop task
 564 for test in desktop laptop ; do
 565     echo > $tmpdir/usr/lib/tasksel/tests/$test &lt;&lt;EOF
 566 #!/bin/sh
 567 exit 2
 568 EOF
 569     chmod a+rx $tmpdir/usr/lib/tasksel/tests/$test
 570 done
 571
 572 DEBIAN_FRONTEND=noninteractive
 573 DEBIAN_PRIORITY=critical
 574 export DEBIAN_FRONTEND DEBIAN_PRIORITY
 575 chroot $tmpdir tasksel --new-install
 576
 577 echo deb $mirror $to main > $tmpdir/etc/apt/sources.list
 578 chroot $tmpdir aptitude update
 579 touch $tmpdir/etc/udev/kernel-upgrade
 580 chroot $tmpdir aptitude -y dist-upgrade
 581 fuser -mv
 582 </pre></blockquote>
 583
 584 <p>I suspect it would be useful to test upgrades with both apt-get and
 585 with aptitude, but I have not had time to look at how they behave
 586 differently so far.  I hope to get a cron job running to do the test
 587 regularly and post the result on the web.  The Gnome upgrade currently
 588 work, while the KDE upgrade fail because of the bug in
 589 kdebase-workspace-data</p>
 590
 591 <p>I am not quite sure what kind of extract from the huge upgrade logs
 592 (KDE 167 KiB, Gnome 516 KiB) it make sense to include in this blog
 593 post, so I will refrain from trying.  I can report that for Gnome,
 594 aptitude report 760 packages upgraded, 448 newly installed, 129 to
 595 remove and 1 not upgraded and 1024MB need to be downloaded while for
 596 KDE the same numbers are 702 packages upgraded, 507 newly installed,
 597 193 to remove and 0 not upgraded and 1117MB need to be downloaded</p>
 598
 599 <p>I am very happy to notice that the Gnome desktop + laptop upgrade
 600 is able to migrate to dependency based boot sequencing and parallel
 601 booting without a hitch.  Was unsure if there were still bugs with
 602 packages failing to clean up their obsolete init.d script during
 603 upgrades, and no such problem seem to affect the Gnome desktop+laptop
 604 packages.</p>
 605
 606       </div>
 607       <div class="tags">
 608
 609
 610         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian edu">debian edu</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 611
 612
 613       </div>
 614     </div>
 615     <div class="padding"></div>
 616
 617     <div class="entry">
 618       <div class="title">
 619         <a href="http://people.skolelinux.org/pere/blog/What_should_start_from__etc_rcS_d__in_Debian____almost_nothing.html">What should start from /etc/rcS.d/ in Debian? - almost nothing</a>
 620       </div>
 621       <div class="date">
 622         30th July 2011
 623       </div>
 624       <div class="body">
 625         <p>In the Debian boot system, several packages include scripts that
 626 are started from /etc/rcS.d/.  In fact, there is a bite more of them
 627 than make sense, and this causes a few problems.  What kind of
 628 problems, you might ask.  There are at least two problems.  The first
 629 is that it is not possible to recover a machine after switching to
 630 runlevel 1.  One need to actually reboot to get the machine back to
 631 the expected state.  The other is that single user boot will sometimes
 632 run into problems because some of the subsystems are activated before
 633 the root login is presented, causing problems when trying to recover a
 634 machine from a problem in that subsystem.  A minor additional point is
 635 that moving more scripts out of rcS.d/ and into the other rc#.d/
 636 directories will increase the amount of scripts that can run in
 637 parallel during boot, and thus decrease the boot time.</p>
 638
 639 <p>So, which scripts should start from rcS.d/.  In short, only the
 640 scripts that _have_ to execute before the root login prompt is
 641 presented during a single user boot should go there.  Everything else
 642 should go into the numeric runlevels.  This means things like
 643 lm-sensors, fuse and x11-common should not run from rcS.d, but from
 644 the numeric runlevels.  Today in Debian, there are around 115 init.d
 645 scripts that are started from rcS.d/, and most of them should be moved
 646 out.  Do your package have one of them?  Please help us make single
 647 user and runlevel 1 better by moving it.</p>
 648
 649 <p>Scripts setting up the screen, keyboard, system partitions
 650 etc. should still be started from rcS.d/, but there is for example no
 651 need to have the network enabled before the single user login prompt
 652 is presented.</p>
 653
 654 <p>As always, things are not so easy to fix as they sound.  To keep
 655 Debian systems working while scripts migrate and during upgrades, the
 656 scripts need to be moved from rcS.d/ to rc2.d/ in reverse dependency
 657 order, ie the scripts that nothing in rcS.d/ depend on can be moved,
 658 and the next ones can only be moved when their dependencies have been
 659 moved first.  This migration must be done sequentially while we ensure
 660 that the package system upgrade packages in the right order to keep
 661 the system state correct.  This will require some coordination when it
 662 comes to network related packages, but most of the packages with
 663 scripts that should migrate do not have anything in rcS.d/ depending
 664 on them.  Some packages have already been updated, like the sudo
 665 package, while others are still left to do.  I wish I had time to work
 666 on this myself, but real live constrains make it unlikely that I will
 667 find time to push this forward.</p>
 668
 669       </div>
 670       <div class="tags">
 671
 672
 673         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 674
 675
 676       </div>
 677     </div>
 678     <div class="padding"></div>
 679
 680     <div class="entry">
 681       <div class="title">
 682         <a href="http://people.skolelinux.org/pere/blog/How_is_booting_into_runlevel_1_different_from_single_user_boots_.html">How is booting into runlevel 1 different from single user boots?</a>
 683       </div>
 684       <div class="date">
 685          4th August 2011
 686       </div>
 687       <div class="body">
 688         <p>Wouter Verhelst have some
 689 <a href="http://grep.be/blog/en/retorts/pere_kubuntu_boot">interesting
 690 comments and opinions</a> on my blog post on
 691 <a href="http://people.skolelinux.org/pere/blog/What_should_start_from__etc_rcS_d__in_Debian____almost_nothing.html">the
 692 need to clean up /etc/rcS.d/ in Debian</a> and my blog post about
 693 <a href="http://people.skolelinux.org/pere/blog/What_is_missing_in_the_Debian_desktop__or_why_my_parents_use_Kubuntu.html">the
 694 default KDE desktop in Debian</a>.  I only have time to address one
 695 small piece of his comment now, and though it best to address the
 696 misunderstanding he bring forward:</p>
 697
 698 <p><blockquote>
 699 Currently, a system admin has four options: [...] boot to a
 700 single-user system (by adding 'single' to the kernel command line;
 701 this runs rcS and rc1 scripts)
 702 </blockquote></p>
 703
 704 <p>This make me believe Wouter believe booting into single user mode
 705 and booting into runlevel 1 is the same.  I am not surprised he
 706 believe this, because it would make sense and is a quite sensible
 707 thing to believe.  But because the boot in Debian is slightly broken,
 708 runlevel 1 do not work properly and it isn't the same as single user
 709 mode.  I'll try to explain what is actually happing, but it is a bit
 710 hard to explain.</p>
 711
 712 <p>Single user mode is defined like this in /etc/inittab:
 713 "<tt>~~:S:wait:/sbin/sulogin</tt>".  This means the only thing that is
 714 executed in single user mode is sulogin.  Single user mode is a boot
 715 state "between" the runlevels, and when booting into single user mode,
 716 only the scripts in /etc/rcS.d/ are executed before the init process
 717 enters the single user state.  When switching to runlevel 1, the state
 718 is in fact not ending in runlevel 1, but it passes through runlevel 1
 719 and end up in the single user mode (see /etc/rc1.d/S03single, which
 720 runs "init -t1 S" to switch to single user mode at the end of runlevel
 721 1.  It is confusing that the 'S' (single user) init mode is not the
 722 mode enabled by /etc/rcS.d/ (which is more like the initial boot
 723 mode).</p>
 724
 725 <p>This summary might make it clearer.  When booting for the first
 726 time into single user mode, the following commands are executed:
 727 "<tt>/etc/init.d/rc S; /sbin/sulogin</tt>".  When booting into
 728 runlevel 1, the following commands are executed: "<tt>/etc/init.d/rc
 729 S; /etc/init.d/rc 1; /sbin/sulogin</tt>".  A problem show up when
 730 trying to continue after visiting single user mode.  Not all services
 731 are started again as they should, causing the machine to end up in an
 732 unpredicatble state.  This is why Debian admins recommend rebooting
 733 after visiting single user mode.</p>
 734
 735 <p>A similar problem with runlevel 1 is caused by the amount of
 736 scripts executed from /etc/rcS.d/.  When switching from say runlevel 2
 737 to runlevel 1, the services started from /etc/rcS.d/ are not properly
 738 stopped when passing through the scripts in /etc/rc1.d/, and not
 739 started again when switching away from runlevel 1 to the runlevels
 740 2-5.  I believe the problem is best fixed by moving all the scripts
 741 out of /etc/rcS.d/ that are not <strong>required</strong> to get a
 742 functioning single user mode during boot.</p>
 743
 744 <p>I have spent several years investigating the Debian boot system,
 745 and discovered this problem a few years ago.  I suspect it originates
 746 from when sysvinit was introduced into Debian, a long time ago.</p>
 747
 748       </div>
 749       <div class="tags">
 750
 751
 752         Tags: <a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem</a>, <a href="http://people.skolelinux.org/pere/blog/tags/debian">debian</a>, <a href="http://people.skolelinux.org/pere/blog/tags/english">english</a>.
 753
 754
 755       </div>
 756     </div>
 757     <div class="padding"></div>
 758
 759     <p style="text-align: right;"><a href="bootsystem.rss"><img src="http://people.skolelinux.org/pere/blog/xml.gif" alt="RSS Feed" width="36" height="14" /></a></p>
 760     <div id="sidebar">
 761
 762
 763
 764 <h2>Archive</h2>
 765 <ul>
 766
 767 <li>2012
 768 <ul>
 769
 770 <li><a href="http://people.skolelinux.org/pere/blog/archive/2012/01/">January (7)</a></li>
 771
 772 <li><a href="http://people.skolelinux.org/pere/blog/archive/2012/02/">February (10)</a></li>
 773
 774 <li><a href="http://people.skolelinux.org/pere/blog/archive/2012/03/">March (17)</a></li>
 775
 776 <li><a href="http://people.skolelinux.org/pere/blog/archive/2012/04/">April (4)</a></li>
 777
 778 </ul></li>
 779
 780 <li>2011
 781 <ul>
 782
 783 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/01/">January (16)</a></li>
 784
 785 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/02/">February (6)</a></li>
 786
 787 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/03/">March (6)</a></li>
 788
 789 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/04/">April (7)</a></li>
 790
 791 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/05/">May (3)</a></li>
 792
 793 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/06/">June (2)</a></li>
 794
 795 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/07/">July (7)</a></li>
 796
 797 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/08/">August (6)</a></li>
 798
 799 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/09/">September (4)</a></li>
 800
 801 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/10/">October (2)</a></li>
 802
 803 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/11/">November (3)</a></li>
 804
 805 <li><a href="http://people.skolelinux.org/pere/blog/archive/2011/12/">December (1)</a></li>
 806
 807 </ul></li>
 808
 809 <li>2010
 810 <ul>
 811
 812 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/01/">January (2)</a></li>
 813
 814 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/02/">February (1)</a></li>
 815
 816 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/03/">March (3)</a></li>
 817
 818 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/04/">April (3)</a></li>
 819
 820 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/05/">May (9)</a></li>
 821
 822 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/06/">June (14)</a></li>
 823
 824 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/07/">July (12)</a></li>
 825
 826 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/08/">August (13)</a></li>
 827
 828 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/09/">September (7)</a></li>
 829
 830 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/10/">October (9)</a></li>
 831
 832 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/11/">November (13)</a></li>
 833
 834 <li><a href="http://people.skolelinux.org/pere/blog/archive/2010/12/">December (12)</a></li>
 835
 836 </ul></li>
 837
 838 <li>2009
 839 <ul>
 840
 841 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/01/">January (8)</a></li>
 842
 843 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/02/">February (8)</a></li>
 844
 845 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/03/">March (12)</a></li>
 846
 847 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/04/">April (10)</a></li>
 848
 849 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/05/">May (9)</a></li>
 850
 851 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/06/">June (3)</a></li>
 852
 853 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/07/">July (4)</a></li>
 854
 855 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/08/">August (3)</a></li>
 856
 857 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/09/">September (1)</a></li>
 858
 859 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/10/">October (2)</a></li>
 860
 861 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/11/">November (3)</a></li>
 862
 863 <li><a href="http://people.skolelinux.org/pere/blog/archive/2009/12/">December (3)</a></li>
 864
 865 </ul></li>
 866
 867 <li>2008
 868 <ul>
 869
 870 <li><a href="http://people.skolelinux.org/pere/blog/archive/2008/11/">November (5)</a></li>
 871
 872 <li><a href="http://people.skolelinux.org/pere/blog/archive/2008/12/">December (7)</a></li>
 873
 874 </ul></li>
 875
 876 </ul>
 877
 878
 879
 880 <h2>Tags</h2>
 881 <ul>
 882
 883  <li><a href="http://people.skolelinux.org/pere/blog/tags/3d-printer">3d-printer (13)</a></li>
 884
 885  <li><a href="http://people.skolelinux.org/pere/blog/tags/amiga">amiga (1)</a></li>
 886
 887  <li><a href="http://people.skolelinux.org/pere/blog/tags/aros">aros (1)</a></li>
 888
 889  <li><a href="http://people.skolelinux.org/pere/blog/tags/bitcoin">bitcoin (2)</a></li>
 890
 891  <li><a href="http://people.skolelinux.org/pere/blog/tags/bootsystem">bootsystem (12)</a></li>
 892
 893  <li><a href="http://people.skolelinux.org/pere/blog/tags/bsa">bsa (2)</a></li>
 894
 895  <li><a href="http://people.skolelinux.org/pere/blog/tags/debian">debian (54)</a></li>
 896
 897  <li><a href="http://people.skolelinux.org/pere/blog/tags/debian edu">debian edu (97)</a></li>
 898
 899  <li><a href="http://people.skolelinux.org/pere/blog/tags/digistan">digistan (7)</a></li>
 900
 901  <li><a href="http://people.skolelinux.org/pere/blog/tags/english">english (121)</a></li>
 902
 903  <li><a href="http://people.skolelinux.org/pere/blog/tags/fiksgatami">fiksgatami (15)</a></li>
 904
 905  <li><a href="http://people.skolelinux.org/pere/blog/tags/fildeling">fildeling (12)</a></li>
 906
 907  <li><a href="http://people.skolelinux.org/pere/blog/tags/intervju">intervju (22)</a></li>
 908
 909  <li><a href="http://people.skolelinux.org/pere/blog/tags/kart">kart (15)</a></li>
 910
 911  <li><a href="http://people.skolelinux.org/pere/blog/tags/ldap">ldap (8)</a></li>
 912
 913  <li><a href="http://people.skolelinux.org/pere/blog/tags/lenker">lenker (4)</a></li>
 914
 915  <li><a href="http://people.skolelinux.org/pere/blog/tags/ltsp">ltsp (1)</a></li>
 916
 917  <li><a href="http://people.skolelinux.org/pere/blog/tags/multimedia">multimedia (14)</a></li>
 918
 919  <li><a href="http://people.skolelinux.org/pere/blog/tags/norsk">norsk (156)</a></li>
 920
 921  <li><a href="http://people.skolelinux.org/pere/blog/tags/nuug">nuug (120)</a></li>
 922
 923  <li><a href="http://people.skolelinux.org/pere/blog/tags/open311">open311 (2)</a></li>
 924
 925  <li><a href="http://people.skolelinux.org/pere/blog/tags/opphavsrett">opphavsrett (24)</a></li>
 926
 927  <li><a href="http://people.skolelinux.org/pere/blog/tags/personvern">personvern (46)</a></li>
 928
 929  <li><a href="http://people.skolelinux.org/pere/blog/tags/raid">raid (1)</a></li>
 930
 931  <li><a href="http://people.skolelinux.org/pere/blog/tags/reprap">reprap (11)</a></li>
 932
 933  <li><a href="http://people.skolelinux.org/pere/blog/tags/rfid">rfid (2)</a></li>
 934
 935  <li><a href="http://people.skolelinux.org/pere/blog/tags/robot">robot (4)</a></li>
 936
 937  <li><a href="http://people.skolelinux.org/pere/blog/tags/rss">rss (1)</a></li>
 938
 939  <li><a href="http://people.skolelinux.org/pere/blog/tags/ruter">ruter (4)</a></li>
 940
 941  <li><a href="http://people.skolelinux.org/pere/blog/tags/sikkerhet">sikkerhet (23)</a></li>
 942
 943  <li><a href="http://people.skolelinux.org/pere/blog/tags/sitesummary">sitesummary (4)</a></li>
 944
 945  <li><a href="http://people.skolelinux.org/pere/blog/tags/standard">standard (24)</a></li>
 946
 947  <li><a href="http://people.skolelinux.org/pere/blog/tags/stavekontroll">stavekontroll (1)</a></li>
 948
 949  <li><a href="http://people.skolelinux.org/pere/blog/tags/stortinget">stortinget (3)</a></li>
 950
 951  <li><a href="http://people.skolelinux.org/pere/blog/tags/surveillance">surveillance (9)</a></li>
 952
 953  <li><a href="http://people.skolelinux.org/pere/blog/tags/valg">valg (6)</a></li>
 954
 955  <li><a href="http://people.skolelinux.org/pere/blog/tags/video">video (23)</a></li>
 956
 957  <li><a href="http://people.skolelinux.org/pere/blog/tags/vitenskap">vitenskap (1)</a></li>
 958
 959  <li><a href="http://people.skolelinux.org/pere/blog/tags/web">web (18)</a></li>
 960
 961 </ul>
 962
 963
 964     </div>
 965     <p style="text-align: right">
 966  Created by <a href="http://steve.org.uk/Software/chronicle">Chronicle v4.4</a>
 967 </p>
 968
 969   </body>
 970 </html>