1 <?xml version=
"1.0" encoding=
"utf-8"?>
2 <rss version='
2.0' xmlns:lj='http://www.livejournal.org/rss/lj/
1.0/' xmlns:
atom=
"http://www.w3.org/2005/Atom">
4 <title>Petter Reinholdtsen
</title>
5 <description></description>
6 <link>http://people.skolelinux.org/pere/blog/
</link>
7 <atom:link href=
"http://people.skolelinux.org/pere/blog/index.rss" rel=
"self" type=
"application/rss+xml" />
10 <title>Circular package dependencies harms apt recovery
</title>
11 <link>http://people.skolelinux.org/pere/blog/Circular_package_dependencies_harms_apt_recovery.html
</link>
12 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/Circular_package_dependencies_harms_apt_recovery.html
</guid>
13 <pubDate>Tue,
27 Jul
2010 23:
50:
00 +
0200</pubDate>
15 <p
>I discovered this while doing
16 <a href=
"http://people.skolelinux.org/pere/blog/Automatic_upgrade_testing_from_Lenny_to_Squeeze.html
">automated
17 testing of upgrades from Debian Lenny to Squeeze
</a
>. A few packages
18 in Debian still got circular dependencies, and it is often claimed
19 that apt and aptitude should be able to handle this just fine, but
20 some times these dependency loops causes apt to fail.
</p
>
22 <p
>An example is from todays
23 <a href=
"http://people.skolelinux.org/~pere/debian-upgrade-testing//test-
20100727-lenny-squeeze-kde-aptitude.txt
">upgrade
24 of KDE using aptitude
</a
>. In it, a bug in kdebase-workspace-data
25 causes perl-modules to fail to upgrade. The cause is simple. If a
26 package fail to unpack, then only part of packages with the circular
27 dependency might end up being unpacked when unpacking aborts, and the
28 ones already unpacked will fail to configure in the recovery phase
29 because its dependencies are unavailable.
</p
>
31 <p
>In this log, the problem manifest itself with this error:
</p
>
33 <blockquote
><pre
>
34 dpkg: dependency problems prevent configuration of perl-modules:
35 perl-modules depends on perl (
>=
5.10.1-
1); however:
36 Version of perl on system is
5.10.0-
19lenny
2.
37 dpkg: error processing perl-modules (--configure):
38 dependency problems - leaving unconfigured
39 </pre
></blockquote
>
41 <p
>The perl/perl-modules circular dependency is already
42 <a href=
"http://bugs.debian.org/
527917">reported as a bug
</a
>, and will
43 hopefully be solved as soon as possible, but it is not the only one,
44 and each one of these loops in the dependency tree can cause similar
45 failures. Of course, they only occur when there are bugs in other
46 packages causing the unpacking to fail, but it is rather nasty when
47 the failure of one package causes the problem to become worse because
48 of dependency loops.
</p
>
51 <a href=
"http://lists.debian.org/debian-devel/
2010/
06/msg00116.html
">the
52 tireless effort by Bill Allombert
</a
>, the number of circular
54 <a href=
"http://debian.semistable.com/debgraph.out.html
">left in Debian
55 is dropping
</a
>, and perhaps it will reach zero one day. :)
</p
>
57 <p
>Todays testing also exposed a bug in
58 <a href=
"http://bugs.debian.org/
590605">update-notifier
</a
> and
59 <a href=
"http://bugs.debian.org/
590604">different behaviour
</a
> between
60 apt-get and aptitude, the latter possibly caused by some circular
61 dependency. Reported both to BTS to try to get someone to look at
67 <title>First Debian Edu test release (alpha0) based on Squeeze is released
</title>
68 <link>http://people.skolelinux.org/pere/blog/First_Debian_Edu_test_release__alpha0__based_on_Squeeze_is_released.html
</link>
69 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/First_Debian_Edu_test_release__alpha0__based_on_Squeeze_is_released.html
</guid>
70 <pubDate>Tue,
27 Jul
2010 17:
45:
00 +
0200</pubDate>
72 <p
>I just posted this announcement culminating several months of work
73 with the next Debian Edu release. Not nearly done, but one major step
77 <p
>This is the first test release based on Squeeze. The focus of this
78 release is to test the user application selection. To have a look,
79 install the standalone profile and let the developers know if the set
80 of installed packages i.e. applications should be modified. If some
81 user application is missing, or if there are some applications that no
82 longer make sense to be included in Debian Edu, please let us know.
83 Also, if a useful application is missing the translation for your
84 language of choice, please let us know too.
</p
>
86 <p
>In addition, feedback and help to polish the desktop (menus,
87 artwork, starters, etc.) is appreciated. We would like to ship a nice
88 and handy KDE4 desktop targeted for schools out of the box.
</p
>
90 <p
>The other profiles should be installable, but there is a lot more
91 work left to be done before they are ready, so do not expect to
94 <p
>Changes compared to the lenny based version
</p
>
97 <li
>Everything from Debian Squeeze
99 <li
>Desktop environment KDE
4.4 =
> the new KDE desktop in
100 combination with some new artwork
101 <li
>Web browser Iceweasel
3.5
102 <li
>OpenOffice.org
3.2
103 <li
>Educational toolbox GCompris
9.3
104 <li
>Music creator Rosegarden
10.04.2
105 <li
>Image editor Gimp
2.6.10
106 <li
>Virtual universe Celestia
1.6.0
107 <li
>Virtual stargazer Stellarium
0.10.4
108 <li
>3D modeler Blender
2.49.2 (new application)
109 <li
>Video editor Kdenlive
0.7.7 (new application)
110 </ul
></li
>
111 <li
>Now using Kerberos for password checking (migration not finished).
117 <li
>SMTP (sender verification)
120 <li
>New experimental roaming workstation profile for laptops.
</li
>
121 <li
>Show welcome page to users when they first log in. The URL is
122 fetched from LDAP.
</li
>
123 <li
>New LXDE desktop option, in addition to KDE (default) and Gnome.
</li
>
124 <li
>General cleanup (not finished)
</li
>
126 <p
>The following features are not working as they should
</p
>
129 <li
>No web based administration tool for creating users and groups. The
130 scripts ldap-createuser-krb and ldap-add-user-to-group can be used
131 for testing.
</li
>
132 <li
>DVD installs are missing debian-installer images for the PXE boot,
133 and do not set up the PXE menu on eth0 because of this. LTSP
134 clients should still boot from eth1 on thin client servers.
</li
>
135 <li
>The restructured KDE menu is not implemented.
</li
>
136 <li
>The LDAP server setup need to be reviewed for security.
</li
>
137 <li
>The LDAP directory structure need to be reworked.
</li
>
138 <li
>Different sets of packages are installed when using the DVD and the
139 netinst CD. More packages are installed using the netinst CD.
</li
>
140 <li
>The jackd package fail to install. This is believed to be caused by
141 some ongoing transition, and hopefully should be solved soon. The
142 jackd1 package can be installed manually for those that need it.
</li
>
143 <li
>Some packages lack translations. See
144 http://wiki.debian.org/DebianEdu/Status/Squeeze for updated status,
145 and help out with translations.
</li
>
148 <p
>To download this multiarch netinstall release you can use
</p
>
151 <li
><a href=
"ftp://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-CD.iso
">ftp://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-CD.iso
</a
></li
>
152 <li
><a href=
"http://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-CD.iso
">http://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-CD.iso
</a
></li
>
153 <li
>rsync -avzP ftp.skolelinux.org::skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-CD.iso
</li
>
155 <p
>To download this multiarch dvd release you can use
</p
>
158 <li
><a href=
"ftp://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-DVD.iso
">ftp://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-DVD.iso
</a
></li
>
159 <li
><a href=
"http://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-DVD.iso
">http://ftp.skolelinux.org/skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-DVD.iso
</a
></li
>
160 <li
>rsync -avzP ftp.skolelinux.org::skolelinux-cd/squeeze-alpha/debian-edu-
6.0.0+edua0-DVD.iso
</li
>
163 <p
>There is no source DVD available yet. It will be prepared when we
164 get closer to the final release.
</p
>
166 <p
>The MD5SUM of these images are
</p
>
169 <li
>3dbf45d59f42a53518b6e3c9ec3b5eb6 debian-edu-
6.0.0+edua0-CD.iso
</li
>
170 <li
>22f2cbfce281d1c6e478be452638675d debian-edu-
6.0.0+edua0-DVD.iso
</li
>
173 <p
>The SHA1SUM of these images are
</p
>
175 <li
>c53d1b69b40cf37cd27aefaf33f6f6a3821bedf0 debian-edu-
6.0.0+edua0-CD.iso
</li
>
176 <li
>2ec29d7db676d59d32197b05c277ffe16348376c debian-edu-
6.0.0+edua0-DVD.iso
</li
>
178 <p
>How to report bugs:
179 http://wiki.debian.org/DebianEdu/HowTo/ReportBugsInBugzilla
</p
>
181 <p
>Please direct replies to debian-edu@lists.debian.org
</p
>
187 <title>One step closer to single signon in Debian Edu
</title>
188 <link>http://people.skolelinux.org/pere/blog/One_step_closer_to_single_signon_in_Debian_Edu.html
</link>
189 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/One_step_closer_to_single_signon_in_Debian_Edu.html
</guid>
190 <pubDate>Sun,
25 Jul
2010 10:
00:
00 +
0200</pubDate>
192 <p
>The last few months me and the other Debian Edu developers have
193 been working hard to get the Debian/Squeeze based version of Debian
194 Edu/Skolelinux into shape. This future version will use Kerberos for
195 authentication, and services are slowly migrated to single signon,
196 getting rid of password questions one at the time.
</p
>
198 <p
>It will also feature a roaming workstation profile with local home
199 directory, for laptops that are only some times on the Skolelinux
200 network, and for this profile a shortcut is created in Gnome and KDE
201 to gain access to the users home directory on the file server. This
202 shortcut uses SMB at the moment, and yesterday I had time to test if
203 SMB mounting had started working in KDE after we added the cifs-utils
204 package. I was pleasantly surprised how well it worked.
</p
>
206 <p
>Thanks to the recent changes to our samba configuration to get it
207 to use Kerberos for authentication, there were no question about user
208 password when mounting the SMB volume. A simple click on the shortcut
209 in the KDE menu, and a window with the home directory popped
212 <p
>One step closer to a single signon solution out of the box in
213 Debian Edu. We already had PAM, LDAP, IMAP and SMTP in place, and now
214 also Samba. Next step is Cups and hopefully also NFS.
</p
>
216 <p
>We had planned a alpha0 release of Debian Edu for today, but thanks
217 to the autobuilder administrators for some architectures being slow to
218 sign packages, we are still missing the fixed LTSP package we need for
219 the release. It was uploaded three days ago with urgency=high, and if
220 it had entered testing yesterday we would have been able to test it in
221 time for a alpha0 release today. As the binaries for ia64 and powerpc
222 still not uploaded to the Debian archive, we need to delay the alpha
223 release another day.
</p
>
225 <p
>If you want to help out with implementing Kerberos for Debian Edu,
226 please contact us on debian-edu@lists.debian.org.
</p
>
231 <title>Digitale restriksjonsmekanismer fikk meg til å slutte å kjøpe musikk
</title>
232 <link>http://people.skolelinux.org/pere/blog/Digitale_restriksjonsmekanismer_fikk_meg_til____slutte____kj__pe_musikk.html
</link>
233 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/Digitale_restriksjonsmekanismer_fikk_meg_til____slutte____kj__pe_musikk.html
</guid>
234 <pubDate>Thu,
22 Jul
2010 23:
50:
00 +
0200</pubDate>
236 <p
>For mange år siden slutte jeg å kjøpe musikk-CDer. Årsaken var at
237 musikkbransjen var godt i gang med å selge platene sine med DRM som
238 gjorde at jeg ikke fikk spilt av musikken jeg kjøpte på utstyret jeg
239 hadde tilgjengelig, dvs. min datamaskin. Det var umulig å se på en
240 plate om den var ødelagt eller ikke, og jeg hadde jo allerede en
241 anseelig samling med plater, så jeg bestemme meg for å slutte å gi
242 penger til en bransje som åpenbart ikke respekterte meg.
</p
>
244 <p
>Jeg har mange titalls dager med musikk på CD i dag. Det meste er
245 lagt i et stort arkiv som kan spilles av fra husets datamaskiner (har
246 ikke rukket rippe alt). Jeg ser dermed ikke behovet for å skaffe mer
247 musikk. De fleste av mine favoritter er i hus, og jeg er dermed godt
250 <p
>Hvis musikkbransjen ønsker mine penger, så må de demonstrere at de
251 setter pris på meg som kunde, og ikke skremme meg bort med DRM og
252 antydninger om at kundene er kriminelle.
</p
>
254 <p
>Filmbransjen er like ille, men mens musikk gjerne varer lenge, er
255 filmer mer ferskvare. Har dermed ikke helt sluttet å kjøpe filmer, men
256 holder meg til DVD-filmer som kan spilles av på mine Linuxbokser.
257 Kommer neppe til å ta i bruk Blueray, og ei heller de nye DRM-greiene
258 «Ultraviolet» som be annonsert her om dagen.
</p
>
263 <title>OpenStreetmap one step closer to having routing on its front page
</title>
264 <link>http://people.skolelinux.org/pere/blog/OpenStreetmap_one_step_closer_to_having_routing_on_its_front_page.html
</link>
265 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/OpenStreetmap_one_step_closer_to_having_routing_on_its_front_page.html
</guid>
266 <pubDate>Sun,
18 Jul
2010 16:
45:
00 +
0200</pubDate>
269 <a href=
"http://feedproxy.google.com/~r/Opengeodata/~
3/wUTCzDZk3lc/project-of-the-week-which-way-home
">todays
270 opengeodata blog entry
</a
>, I just discovered that the
271 OpenStreetmap.org site have gotten
272 <a href=
"http://nroets.dev.openstreetmap.org/demo/index.html?layers=B000FTFTT
">support
273 for calculating routes
</a
>. The support is still experimental and
274 only available from the development server, until more experience is
275 gathered on the user interface and any scalability issues.
</p
>
277 <p
>Earlier, the routing I knew about using the OpenStreetmap.org data
278 was provided by
<a href=
"http://maps.cloudmade.com/
">Cloudmade
</a
>,
279 but having it on the main page is required to make everyone aware of
280 the issue. I
've had people reject Openstreetmap.org as a viable
281 alternative for them because the front page lacked routing support,
282 and I hope their needs will be catered for when routing show up on the
283 www.openstreetmap.org front page.
</p
>
288 <title>What are they searching for - PowerDNS and ISC DHCP in LDAP
</title>
289 <link>http://people.skolelinux.org/pere/blog/What_are_they_searching_for___PowerDNS_and_ISC_DHCP_in_LDAP.html
</link>
290 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/What_are_they_searching_for___PowerDNS_and_ISC_DHCP_in_LDAP.html
</guid>
291 <pubDate>Sat,
17 Jul
2010 21:
00:
00 +
0200</pubDate>
294 <a href=
"http://people.skolelinux.org/pere/blog/Time_for_new__LDAP_schemas_replacing_RFC_2307_.html
">followup
</a
>
296 <a href=
"http://people.skolelinux.org/pere/blog/Idea_for_a_change_to_LDAP_schemas_allowing_DNS_and_DHCP_info_to_be_combined_into_one_object.html
">previous
298 <a href=
"http://people.skolelinux.org/pere/blog/Combining_PowerDNS_and_ISC_DHCP_LDAP_objects.html
">merging
299 all
</a
> the computer related LDAP objects in Debian Edu.
</p
>
301 <p
>As a step to try to see if it possible to merge the DNS and DHCP
302 LDAP objects, I have had a look at how the packages pdns-backend-ldap
303 and dhcp3-server-ldap in Debian use the LDAP server. The two
304 implementations are quite different in how they use LDAP.
</p
>
306 To get this information, I started slapd with debugging enabled and
307 dumped the debug output to a file to get the LDAP searches performed
308 on a Debian Edu main-server. Here is a summary.
310 <p
><strong
>powerdns
</strong
></p
>
312 <a href=
"http://www.linuxnetworks.de/doc/index.php/PowerDNS_LDAP_Backend
">Clues
313 on how to
</a
> set up PowerDNS to use a LDAP backend is available on
316 <p
>PowerDNS have two modes of operation using LDAP as its backend.
317 One
"strict
" mode where the forward and reverse DNS lookups are done
318 using the same LDAP objects, and a
"tree
" mode where the forward and
319 reverse entries are in two different subtrees in LDAP with a structure
320 based on the DNS names, as in tjener.intern and
321 2.2.0.10.in-addr.arpa.
</p
>
323 <p
>In tree mode, the server is set up to use a LDAP subtree as its
324 base, and uses a
"base
" scoped search for the DNS name by adding
325 "dc=tjener,dc=intern,
" to the base with a filter for
326 "(associateddomain=tjener.intern)
" for the forward entry and
327 "dc=
2,dc=
2,dc=
0,dc=
10,dc=in-addr,dc=arpa,
" with a filter for
328 "(associateddomain=
2.2.0.10.in-addr.arpa)
" for the reverse entry. For
329 forward entries, it is looking for attributes named dnsttl, arecord,
330 nsrecord, cnamerecord, soarecord, ptrrecord, hinforecord, mxrecord,
331 txtrecord, rprecord, afsdbrecord, keyrecord, aaaarecord, locrecord,
332 srvrecord, naptrrecord, kxrecord, certrecord, dsrecord, sshfprecord,
333 ipseckeyrecord, rrsigrecord, nsecrecord, dnskeyrecord, dhcidrecord,
334 spfrecord and modifytimestamp. For reverse entries it is looking for
335 the attributes dnsttl, arecord, nsrecord, cnamerecord, soarecord,
336 ptrrecord, hinforecord, mxrecord, txtrecord, rprecord, aaaarecord,
337 locrecord, srvrecord, naptrrecord and modifytimestamp. The equivalent
338 ldapsearch commands could look like this:
</p
>
340 <blockquote
><pre
>
342 -b dc=tjener,dc=intern,ou=hosts,dc=skole,dc=skolelinux,dc=no \
343 -s base -x
'(associateddomain=tjener.intern)
' dNSTTL aRecord nSRecord \
344 cNAMERecord sOARecord pTRRecord hInfoRecord mXRecord tXTRecord \
345 rPRecord aFSDBRecord KeyRecord aAAARecord lOCRecord sRVRecord \
346 nAPTRRecord kXRecord certRecord dSRecord sSHFPRecord iPSecKeyRecord \
347 rRSIGRecord nSECRecord dNSKeyRecord dHCIDRecord sPFRecord modifyTimestamp
350 -b dc=
2,dc=
2,dc=
0,dc=
10,dc=in-addr,dc=arpa,ou=hosts,dc=skole,dc=skolelinux,dc=no \
351 -s base -x
'(associateddomain=
2.2.0.10.in-addr.arpa)
'
352 dnsttl, arecord, nsrecord, cnamerecord soarecord ptrrecord \
353 hinforecord mxrecord txtrecord rprecord aaaarecord locrecord \
354 srvrecord naptrrecord modifytimestamp
355 </pre
></blockquote
>
357 <p
>In Debian Edu/Lenny, the PowerDNS tree mode is used with
358 ou=hosts,dc=skole,dc=skolelinux,dc=no as the base, and these are two
359 example LDAP objects used there. In addition to these objects, the
360 parent objects all th way up to ou=hosts,dc=skole,dc=skolelinux,dc=no
361 also exist.
</p
>
363 <blockquote
><pre
>
364 dn: dc=tjener,dc=intern,ou=hosts,dc=skole,dc=skolelinux,dc=no
366 objectclass: dnsdomain
367 objectclass: domainrelatedobject
370 associateddomain: tjener.intern
372 dn: dc=
2,dc=
2,dc=
0,dc=
10,dc=in-addr,dc=arpa,ou=hosts,dc=skole,dc=skolelinux,dc=no
374 objectclass: dnsdomain2
375 objectclass: domainrelatedobject
377 ptrrecord: tjener.intern
378 associateddomain:
2.2.0.10.in-addr.arpa
379 </pre
></blockquote
>
381 <p
>In strict mode, the server behaves differently. When looking for
382 forward DNS entries, it is doing a
"subtree
" scoped search with the
383 same base as in the tree mode for a object with filter
384 "(associateddomain=tjener.intern)
" and requests the attributes dnsttl,
385 arecord, nsrecord, cnamerecord, soarecord, ptrrecord, hinforecord,
386 mxrecord, txtrecord, rprecord, aaaarecord, locrecord, srvrecord,
387 naptrrecord and modifytimestamp. For reverse entires it also do a
388 subtree scoped search but this time the filter is
"(arecord=
10.0.2.2)
"
389 and the requested attributes are associateddomain, dnsttl and
390 modifytimestamp. In short, in strict mode the objects with ptrrecord
391 go away, and the arecord attribute in the forward object is used
394 <p
>The forward and reverse searches can be simulated using ldapsearch
397 <blockquote
><pre
>
398 ldapsearch -h ldap -b ou=hosts,dc=skole,dc=skolelinux,dc=no -s sub -x \
399 '(associateddomain=tjener.intern)
' dNSTTL aRecord nSRecord \
400 cNAMERecord sOARecord pTRRecord hInfoRecord mXRecord tXTRecord \
401 rPRecord aFSDBRecord KeyRecord aAAARecord lOCRecord sRVRecord \
402 nAPTRRecord kXRecord certRecord dSRecord sSHFPRecord iPSecKeyRecord \
403 rRSIGRecord nSECRecord dNSKeyRecord dHCIDRecord sPFRecord modifyTimestamp
405 ldapsearch -h ldap -b ou=hosts,dc=skole,dc=skolelinux,dc=no -s sub -x \
406 '(arecord=
10.0.2.2)
' associateddomain dnsttl modifytimestamp
407 </pre
></blockquote
>
409 <p
>In addition to the forward and reverse searches , there is also a
410 search for SOA records, which behave similar to the forward and
411 reverse lookups.
</p
>
413 <p
>A thing to note with the PowerDNS behaviour is that it do not
414 specify any objectclass names, and instead look for the attributes it
415 need to generate a DNS reply. This make it able to work with any
416 objectclass that provide the needed attributes.
</p
>
418 <p
>The attributes are normally provided in the cosine (RFC
1274) and
419 dnsdomain2 schemas. The latter is used for reverse entries like
420 ptrrecord and recent DNS additions like aaaarecord and srvrecord.
</p
>
422 <p
>In Debian Edu, we have created DNS objects using the object classes
423 dcobject (for dc), dnsdomain or dnsdomain2 (structural, for the DNS
424 attributes) and domainrelatedobject (for associatedDomain). The use
425 of structural object classes make it impossible to combine these
426 classes with the object classes used by DHCP.
</p
>
428 <p
>There are other schemas that could be used too, for example the
429 dnszone structural object class used by Gosa and bind-sdb for the DNS
430 attributes combined with the domainrelatedobject object class, but in
431 this case some unused attributes would have to be included as well
432 (zonename and relativedomainname).
</p
>
434 <p
>My proposal for Debian Edu would be to switch PowerDNS to strict
435 mode and not use any of the existing objectclasses (dnsdomain,
436 dnsdomain2 and dnszone) when one want to combine the DNS information
437 with DHCP information, and instead create a auxiliary object class
438 defined something like this (using the attributes defined for
439 dnsdomain and dnsdomain2 or dnszone):
</p
>
441 <blockquote
><pre
>
442 objectclass ( some-oid NAME
'dnsDomainAux
'
445 MAY ( ARecord $ MDRecord $ MXRecord $ NSRecord $ SOARecord $ CNAMERecord $
446 DNSTTL $ DNSClass $ PTRRecord $ HINFORecord $ MINFORecord $
447 TXTRecord $ SIGRecord $ KEYRecord $ AAAARecord $ LOCRecord $
448 NXTRecord $ SRVRecord $ NAPTRRecord $ KXRecord $ CERTRecord $
449 A6Record $ DNAMERecord
451 </pre
></blockquote
>
453 <p
>This will allow any object to become a DNS entry when combined with
454 the domainrelatedobject object class, and allow any entity to include
455 all the attributes PowerDNS wants. I
've sent an email to the PowerDNS
456 developers asking for their view on this schema and if they are
457 interested in providing such schema with PowerDNS, and I hope my
458 message will be accepted into their mailing list soon.
</p
>
460 <p
><strong
>ISC dhcp
</strong
></p
>
462 <p
>The DHCP server searches for specific objectclass and requests all
463 the object attributes, and then uses the attributes it want. This
464 make it harder to figure out exactly what attributes are used, but
465 thanks to the working example in Debian Edu I can at least get an idea
466 what is needed without having to read the source code.
</p
>
468 <p
>In the DHCP server configuration, the LDAP base to use and the
469 search filter to use to locate the correct dhcpServer entity is
470 stored. These are the relevant entries from
471 /etc/dhcp3/dhcpd.conf:
</p
>
473 <blockquote
><pre
>
474 ldap-base-dn
"dc=skole,dc=skolelinux,dc=no
";
475 ldap-dhcp-server-cn
"dhcp
";
476 </pre
></blockquote
>
478 <p
>The DHCP server uses this information to nest all the DHCP
479 configuration it need. The cn
"dhcp
" is located using the given LDAP
480 base and the filter
"(
&(objectClass=dhcpServer)(cn=dhcp))
". The
481 search result is this entry:
</p
>
483 <blockquote
><pre
>
484 dn: cn=dhcp,dc=skole,dc=skolelinux,dc=no
487 objectClass: dhcpServer
488 dhcpServiceDN: cn=DHCP Config,dc=skole,dc=skolelinux,dc=no
489 </pre
></blockquote
>
491 <p
>The content of the dhcpServiceDN attribute is next used to locate the
492 subtree with DHCP configuration. The DHCP configuration subtree base
493 is located using a base scope search with base
"cn=DHCP
494 Config,dc=skole,dc=skolelinux,dc=no
" and filter
495 "(
&(objectClass=dhcpService)(|(dhcpPrimaryDN=cn=dhcp,dc=skole,dc=skolelinux,dc=no)(dhcpSecondaryDN=cn=dhcp,dc=skole,dc=skolelinux,dc=no)))
".
496 The search result is this entry:
</p
>
498 <blockquote
><pre
>
499 dn: cn=DHCP Config,dc=skole,dc=skolelinux,dc=no
502 objectClass: dhcpService
503 objectClass: dhcpOptions
504 dhcpPrimaryDN: cn=dhcp, dc=skole,dc=skolelinux,dc=no
505 dhcpStatements: ddns-update-style none
506 dhcpStatements: authoritative
507 dhcpOption: smtp-server code
69 = array of ip-address
508 dhcpOption: www-server code
72 = array of ip-address
509 dhcpOption: wpad-url code
252 = text
510 </pre
></blockquote
>
512 <p
>Next, the entire subtree is processed, one level at the time. When
513 all the DHCP configuration is loaded, it is ready to receive requests.
514 The subtree in Debian Edu contain objects with object classes
515 top/dhcpService/dhcpOptions, top/dhcpSharedNetwork/dhcpOptions,
516 top/dhcpSubnet, top/dhcpGroup and top/dhcpHost. These provide options
517 and information about netmasks, dynamic range etc. Leaving out the
518 details here because it is not relevant for the focus of my
519 investigation, which is to see if it is possible to merge dns and dhcp
520 related computer objects.
</p
>
522 <p
>When a DHCP request come in, LDAP is searched for the MAC address
523 of the client (
00:
00:
00:
00:
00:
00 in this example), using a subtree
524 scoped search with
"cn=DHCP Config,dc=skole,dc=skolelinux,dc=no
" as
525 the base and
"(
&(objectClass=dhcpHost)(dhcpHWAddress=ethernet
526 00:
00:
00:
00:
00:
00))
" as the filter. This is what a host object look
529 <blockquote
><pre
>
530 dn: cn=hostname,cn=group1,cn=THINCLIENTS,cn=DHCP Config,dc=skole,dc=skolelinux,dc=no
533 objectClass: dhcpHost
534 dhcpHWAddress: ethernet
00:
00:
00:
00:
00:
00
535 dhcpStatements: fixed-address hostname
536 </pre
></blockquote
>
538 <p
>There is less flexiblity in the way LDAP searches are done here.
539 The object classes need to have fixed names, and the configuration
540 need to be stored in a fairly specific LDAP structure. On the
541 positive side, the invidiual dhcpHost entires can be anywhere without
542 the DN pointed to by the dhcpServer entries. The latter should make
543 it possible to group all host entries in a subtree next to the
544 configuration entries, and this subtree can also be shared with the
545 DNS server if the schema proposed above is combined with the dhcpHost
546 structural object class.
548 <p
><strong
>Conclusion
</strong
></p
>
550 <p
>The PowerDNS implementation seem to be very flexible when it come
551 to which LDAP schemas to use. While its
"tree
" mode is rigid when it
552 come to the the LDAP structure, the
"strict
" mode is very flexible,
553 allowing DNS objects to be stored anywhere under the base cn specified
554 in the configuration.
</p
>
556 <p
>The DHCP implementation on the other hand is very inflexible, both
557 regarding which LDAP schemas to use and which LDAP structure to use.
558 I guess one could implement ones own schema, as long as the
559 objectclasses and attributes have the names used, but this do not
560 really help when the DHCP subtree need to have a fairly fixed
563 <p
>Based on the observed behaviour, I suspect a LDAP structure like
564 this might work for Debian Edu:
</p
>
566 <blockquote
><pre
>
568 cn=machine-info (dhcpService) - dhcpServiceDN points here
570 cn=dhcp-internal (dhcpSharedNetwork/dhcpOptions)
571 cn=
10.0.2.0 (dhcpSubnet)
572 cn=group1 (dhcpGroup/dhcpOptions)
573 cn=dhcp-thinclients (dhcpSharedNetwork/dhcpOptions)
574 cn=
192.168.0.0 (dhcpSubnet)
575 cn=group1 (dhcpGroup/dhcpOptions)
576 ou=machines - PowerDNS base points here
577 cn=hostname (dhcpHost/domainrelatedobject/dnsDomainAux)
578 </pre
></blockquote
>
580 <P
>This is not tested yet. If the DHCP server require the dhcpHost
581 entries to be in the dhcpGroup subtrees, the entries can be stored
582 there instead of a common machines subtree, and the PowerDNS base
583 would have to be moved one level up to the machine-info subtree.
</p
>
585 <p
>The combined object under the machines subtree would look something
588 <blockquote
><pre
>
589 dn: dc=hostname,ou=machines,cn=machine-info,dc=skole,dc=skolelinux,dc=no
592 objectClass: dhcpHost
593 objectclass: domainrelatedobject
594 objectclass: dnsDomainAux
595 associateddomain: hostname.intern
597 dhcpHWAddress: ethernet
00:
00:
00:
00:
00:
00
598 dhcpStatements: fixed-address hostname.intern
599 </pre
></blockquote
>
601 </p
>One could even add the LTSP configuration associated with a given
602 machine, as long as the required attributes are available in a
603 auxiliary object class.
</p
>
608 <title>Combining PowerDNS and ISC DHCP LDAP objects
</title>
609 <link>http://people.skolelinux.org/pere/blog/Combining_PowerDNS_and_ISC_DHCP_LDAP_objects.html
</link>
610 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/Combining_PowerDNS_and_ISC_DHCP_LDAP_objects.html
</guid>
611 <pubDate>Wed,
14 Jul
2010 23:
45:
00 +
0200</pubDate>
613 <p
>For a while now, I have wanted to find a way to change the DNS and
614 DHCP services in Debian Edu to use the same LDAP objects for a given
615 computer, to avoid the possibility of having a inconsistent state for
616 a computer in LDAP (as in DHCP but no DNS entry or the other way
617 around) and make it easier to add computers to LDAP.
</p
>
619 <p
>I
've looked at how powerdns and dhcpd is using LDAP, and using this
620 information finally found a solution that seem to work.
</p
>
622 <p
>The old setup required three LDAP objects for a given computer.
623 One forward DNS entry, one reverse DNS entry and one DHCP entry. If
624 we switch powerdns to use its strict LDAP method (ldap-method=strict
625 in pdns-debian-edu.conf), the forward and reverse DNS entries are
626 merged into one while making it impossible to transfer the reverse map
627 to a slave DNS server.
</p
>
629 <p
>If we also replace the object class used to get the DNS related
630 attributes to one allowing these attributes to be combined with the
631 dhcphost object class, we can merge the DNS and DHCP entries into one.
632 I
've written such object class in the dnsdomainaux.schema file (need
633 proper OIDs, but that is a minor issue), and tested the setup. It
634 seem to work.
</p
>
636 <p
>With this test setup in place, we can get away with one LDAP object
637 for both DNS and DHCP, and even the LTSP configuration I suggested in
638 an earlier email. The combined LDAP object will look something like
641 <blockquote
><pre
>
642 dn: cn=hostname,cn=group1,cn=THINCLIENTS,cn=DHCP Config,dc=skole,dc=skolelinux,dc=no
644 objectClass: dhcphost
645 objectclass: domainrelatedobject
646 objectclass: dnsdomainaux
647 associateddomain: hostname.intern
649 dhcphwaddress: ethernet
00:
00:
00:
00:
00:
00
650 dhcpstatements: fixed-address hostname
652 </pre
></blockquote
>
654 <p
>The DNS server uses the associateddomain and arecord entries, while
655 the DHCP server uses the dhcphwaddress and dhcpstatements entries
656 before asking DNS to resolve the fixed-adddress. LTSP will use
657 dhcphwaddress or associateddomain and the ldapconfig* attributes.
</p
>
659 <p
>I am not yet sure if I can get the DHCP server to look for its
660 dhcphost in a different location, to allow us to put the objects
661 outside the
"DHCP Config
" subtree, but hope to figure out a way to do
662 that. If I can
't figure out a way to do that, we can still get rid of
663 the hosts subtree and move all its content into the DHCP Config tree
664 (which probably should be renamed to be more related to the new
665 content. I suspect cn=dnsdhcp,ou=services or something like that
666 might be a good place to put it.
</p
>
668 <p
>If you want to help out with implementing this for Debian Edu,
669 please contact us on debian-edu@lists.debian.org.
</p
>
674 <title>Idea for storing LTSP configuration in LDAP
</title>
675 <link>http://people.skolelinux.org/pere/blog/Idea_for_storing_LTSP_configuration_in_LDAP.html
</link>
676 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/Idea_for_storing_LTSP_configuration_in_LDAP.html
</guid>
677 <pubDate>Sun,
11 Jul
2010 22:
00:
00 +
0200</pubDate>
679 <p
>Vagrant mentioned on IRC today that ltsp_config now support
680 sourcing files from /usr/share/ltsp/ltsp_config.d/ on the thin
681 clients, and that this can be used to fetch configuration from LDAP if
682 Debian Edu choose to store configuration there.
</p
>
684 <p
>Armed with this information, I got inspired and wrote a test module
685 to get configuration from LDAP. The idea is to look up the MAC
686 address of the client in LDAP, and look for attributes on the form
687 ltspconfigsetting=value, and use this to export SETTING=value to the
688 LTSP clients.
</p
>
690 <p
>The goal is to be able to store the LTSP configuration attributes
691 in a
"computer
" LDAP object used by both DNS and DHCP, and thus
692 allowing us to store all information about a computer in one place.
</p
>
694 <p
>This is a untested draft implementation, and I welcome feedback on
695 this approach. A real LDAP schema for the ltspClientAux objectclass
696 need to be written. Comments, suggestions, etc?
</p
>
698 <blockquote
><pre
>
699 # Store in /opt/ltsp/$arch/usr/share/ltsp/ltsp_config.d/ldap-config
701 # Fetch LTSP client settings from LDAP based on MAC address
703 # Uses ethernet address as stored in the dhcpHost objectclass using
704 # the dhcpHWAddress attribute or ethernet address stored in the
705 # ieee802Device objectclass with the macAddress attribute.
707 # This module is written to be schema agnostic, and only depend on the
708 # existence of attribute names.
710 # The LTSP configuration variables are saved directly using a
711 # ltspConfig prefix and uppercasing the rest of the attribute name.
712 # To set the SERVER variable, set the ltspConfigServer attribute.
714 # Some LDAP schema should be created with all the relevant
715 # configuration settings. Something like this should work:
717 # objectclass (
1.1.2.2 NAME
'ltspClientAux
'
720 # MAY ( ltspConfigServer $ ltsConfigSound $ ... )
722 LDAPSERVER=$(debian-edu-ldapserver)
723 if [
"$LDAPSERVER
" ] ; then
724 LDAPBASE=$(debian-edu-ldapserver -b)
725 for MAC in $(LANG=C ifconfig |grep -i hwaddr| awk
'{print $
5}
'|sort -u) ; do
726 filter=
"(|(dhcpHWAddress=ethernet $MAC)(macAddress=$MAC))
"
727 ldapsearch -h
"$LDAPSERVER
" -b
"$LDAPBASE
" -v -x
"$filter
" | \
728 grep
'^ltspConfig
' | while read attr value ; do
729 # Remove prefix and convert to upper case
730 attr=$(echo $attr | sed
's/^ltspConfig//i
' | tr a-z A-Z)
731 # bass value on to clients
732 eval
"$attr=$value; export $attr
"
736 </pre
></blockquote
>
738 <p
>I
'm not sure this shell construction will work, because I suspect
739 the while block might end up in a subshell causing the variables set
740 there to not show up in ltsp-config, but if that is the case I am sure
741 the code can be restructured to make sure the variables are passed on.
742 I expect that can be solved with some testing. :)
</p
>
744 <p
>If you want to help out with implementing this for Debian Edu,
745 please contact us on debian-edu@lists.debian.org.
</p
>
747 <p
>Update
2010-
07-
17: I am aware of another effort to store LTSP
748 configuration in LDAP that was created around year
2000 by
749 <a href=
"http://www.pcxperience.com/thinclient/documentation/ldap.html
">PC
750 Xperience, Inc.,
2000</a
>. I found its
751 <a href=
"http://people.redhat.com/alikins/ltsp/ldap/
">files
</a
> on a
752 personal home page over at redhat.com.
</p
>
757 <title>jXplorer, a very nice LDAP GUI
</title>
758 <link>http://people.skolelinux.org/pere/blog/jXplorer__a_very_nice_LDAP_GUI.html
</link>
759 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/jXplorer__a_very_nice_LDAP_GUI.html
</guid>
760 <pubDate>Fri,
9 Jul
2010 12:
55:
00 +
0200</pubDate>
763 <a href=
"http://people.skolelinux.org/pere/blog/LUMA__a_very_nice_LDAP_GUI.html
">my
764 last post
</a
> about available LDAP tools in Debian, I was told about a
765 LDAP GUI that is even better than luma. The java application
766 <a href=
"http://jxplorer.org/
">jXplorer
</a
> is claimed to be capable of
767 moving LDAP objects and subtrees using drag-and-drop, and can
768 authenticate using Kerberos. I have only tested the Kerberos
769 authentication, but do not have a LDAP setup allowing me to rewrite
770 LDAP with my test user yet. It is
771 <a href=
"http://packages.qa.debian.org/j/jxplorer.html
">available in
772 Debian
</a
> testing and unstable at the moment. The only problem I
773 have with it is how it handle errors. If something go wrong, its
774 non-intuitive behaviour require me to go through some query work list
775 and remove the failing query. Nothing big, but very annoying.
</p
>
780 <title>MS Word krøller det til for politiet?
</title>
781 <link>http://people.skolelinux.org/pere/blog/MS_Word_kr__ller_det_til_for_politiet_.html
</link>
782 <guid isPermaLink=
"true">http://people.skolelinux.org/pere/blog/MS_Word_kr__ller_det_til_for_politiet_.html
</guid>
783 <pubDate>Thu,
8 Jul
2010 14:
00:
00 +
0200</pubDate>
785 <p
>De siste dagene har Aftenposten
786 <a href=
"http://www.aftenposten.no/nyheter/iriks/article3718597.ece
">fortalt
</a
>
787 <a href=
"http://www.aftenposten.no/nyheter/iriks/article3724249.ece
">hvordan
</a
>
788 politet har brukt skriveverktøy som ikke håndterer arabisk tekst og
789 tekst som skal skrives fra høyre mot venstre når de har laget
790 løpeseddel for å be om informasjon fra publikum. Resultatet har vært
791 en uleselig arabisk-bit på løpeseddelen. Feilen har oppstått når
792 teksten har blitt
"kopiert inn i programvare som ikke har støtte for
793 språk som skrives fra høyre mot venstre
", og jeg er ganske sikker på
794 at det er snakk om Microsoft Office i dette tilfellet. Er det slik at
795 MS Office i norsk språkdrakt ikke har støtte for tekst som skal
796 skrives fra høyre mot venstre? Jeg tror alle utgaver av
797 OpenOffice.org har slik støtte, og det er jo ikke veldig vanskelig å
798 la slik støtte finnes i alle utgaver av et program hvis støtten først
799 er utviklet. Aftenpostens melding får meg til å undre om problemet
800 ville vært unngått hvis politiet brukte OpenOffice.org i stedet for MS
803 <p
>Mon tro om det er flere eksempler på at MS Office har ødelagt for
804 offentlig myndighet?
</p
>