source: branches/fc11-dev/server/doc/install-howto.sh @ 1211

Last change on this file since 1211 was 1211, checked in by ezyang, 15 years ago
More install information.
File size: 15.0 KB
RevLine 
[1058]1# This document is a how-to for installing a Fedora scripts.mit.edu server.
[181]2
[1063]3set -e -x
4
5[ -e /scripts-boot-count ] || echo 0 > /scripts-boot-count
6
7source_server="old-faithful.mit.edu"
8
9boot=${1:$(cat /scripts-boot-count)}
10
[1205]11branch=branches/fc11-dev
12
[1063]13doreboot() {
14    echo $(( $boot + 1 )) > /scripts-boot-count;
15    shutdown -r now "Rebooting for step $(cat /scripts-boot-count)"
16}
17
18YUM() {
19    NSS_NONLOCAL_IGNORE=1 yum "$@"
20}
21
[1058]22# Helper files for the install are located in server/fedora/config.
[861]23
[1058]24# Start with a normal install of Fedora.
[861]25
[1063]26if [ $boot = 0 ]; then
[1058]27# When the initial configuration screen comes up, under "Firewall
28# configuration", disable the firewall, and under "System services", leave
29# enabled (as of Fedora 9) acpid, anacron, atd, cpuspeed, crond,
30# firstboot, fuse, haldaemon, ip6tables, iptables, irqbalance,
31# kerneloops, mdmonitor, messagebus, microcode_ctl, netfs, network, nscd, ntpd,
32# sshd, udev-post, and nothing else.
[1063]33    echo "--disabled" > /etc/sysconfig/system-config-firewall
34    for i in NetworkManager avahi-daemon bluetooth cups isdn nfslock pcscd restorecond rpcbind rpcgssd rpcidmapd sendmail; do
35        chkconfig "$i" off
36    done
[1056]37
[1058]38# Edit /etc/selinux/config so it has SELINUX=disabled and reboot.
[1063]39    sed -i 's/^SELINUX=.*/SELINUX=disabled/' /etc/selinux/config
40    doreboot
41fi
[861]42
[1063]43if [ $boot = 1 ]; then
44# Create a scripts-build user account, and set up rpm to build in
45# $HOME by doing a
46# cp config/home/scripts-build/.rpmmacros /home/scripts-build/
47# (If you just use the default setup, it will generate packages
48# in /usr/src/redhat.)
49    adduser scripts-build
50
[1058]51# Check out the scripts.mit.edu svn repository. Configure svn not to cache
52# credentials.
[387]53
[1063]54    YUM install -y subversion
55
56    cd /srv
[1205]57    svn co svn://$source_server/$branch repository
[1063]58
59    sed -i 's/^(# *)*store-passwords.*/store-passwords = no/' /root/.subversion/config
60    sed -i 's/^(# *)*store-auth-creds.*/store-auth-creds = no/' /root/.subversion/config
61
62    chown -R scripts-build /srv/repository
63
[1058]64# cd to server/fedora in the svn repository.
[1063]65    cd /srv/repository/server/fedora
[387]66
[1058]67# Run "make install-deps" to install various prereqs.  Nonstandard
68# deps are in /mit/scripts/rpm.
[1205]69    YUM install -y make
[1063]70    make install-deps
[387]71
[1063]72# Install bind
73    YUM install -y bind
[934]74
[1063]75# Check out the scripts /etc configuration
76    cd /root
[1205]77    svn co svn://scripts.mit.edu/$branch/server/fedora/config/etc etc
78    # backslash to make us not use the alias
[1063]79    \cp -a etc /
[785]80
[1205]81# NOTE: You will have just lost DNS resolution and the abilit
82# to do password SSH in
83
84    service named start
85    chkconfig named on
86
87# XXX: This sometimes doesn't exist, but it really sucks if it
88# does exist. So check for it.
[1179]89# yum remove nss_ldap, because nss-ldapd conflicts with it
90
[1205]91# In the case of the Kerberos libraries, you'll be told that
92# there are conflicting files with the 64-bit versions of the packages,
93# which we scriptsify.  You'll have to use --force to install those
94# rpms despite the conflicts.  After doing that, you may want to
95# install the corresponding 64-bit scriptsified versions again, just
96# to be safe in case the 32-bit versions overwrite files that differ.
97# When you try this, it will complain that you already have the same
98# version installed; again, you'll need to use --force to do it anyway.
99
100# We need yumdownloader to force some RPMs
101    # XXX: This might be wrong. Sanity check what packages ou
102    # have when done
103    YUM install -y yum-utils
104    yumdownloader krb5-libs
105    # XXX: These version numbers are hardcoded, need some cli-fu to generalize
106    rpm -i krb5-libs-1.6.3-20.fc11.i586.rpm
107    rpm -U --force krb5-libs-1.6.3-20.fc11.scripts.1138.x86_64.rpm
108
[1058]109# env NSS_NONLOCAL_IGNORE=1 yum install scripts-base
[1063]110    YUM install -y scripts-base
[387]111
[1205]112# Install mit-zephyr
113    YUM install -y mit-zephyr
[796]114
[1058]115# Remember to set NSS_NONLOCAL_IGNORE=1 anytime you're setting up
116# anything, e.g. using yum. Otherwise useradd will query LDAP in a stupid way
[1205]117# that makes it hang forever. (This is why we're using YUM, not yum)
[881]118
[1058]119# Reload the iptables config to take down the restrictive firewall
[1205]120    service iptables restart
[862]121
[1058]122# Copy over root's dotfiles from one of the other machines.
[1205]123# Perhaps a useful change is to remove the default aliases
[803]124
[1058]125# Replace rsyslog with syslog-ng by doing:
[1205]126    rpm -e --nodeps rsyslog
127    YUM install -y syslog-ng
[1211]128    chkconfig syslog-ng on
[861]129
[1205]130# Install various dependencies of the scripts system, including
131# glibc-devel.i586 (ezyang: already installed for me),
132# python-twisted-core (ditto), mod_fcgid, nrpe, nagios-plugins-all.
133    YUM install -y mod_fcgid
134    YUM install -y nrpe
135    YUM install -y nagios-plugins-all
[788]136
[1058]137# Disable NetworkManager with chkconfig NetworkManager off. Configure
138# networking on the front end and back end, and the routing table to send
139# traffic over the back end. Make sure that chkconfig reports "network" on, so
140# that the network will still be configured at next boot.
[1205]141# ezyang: For me, NetworkManager was not installed at this point, and
142# we had already done the basic config for networking front end and
143# back end (because I wanted ssh access, and not just conserver access)
[788]144
[1058]145# Fix the openafs /usr/vice/etc <-> /etc/openafs mapping by changing
146#  /usr/vice/etc/cacheinfo to contain:
147#       /afs:/usr/vice/cache:10000000
[1179]148# Also fix ThisCell to contain athena.mit.edu in both directories
[1205]149    echo "/afs:/usr/vice/cache:10000000" > /usr/vice/etc/cacheinfo
150    # ezyang: ThisCell on b-k and c-w don't have anything special
151    # written here
[788]152
[1058]153# Figure out why Zephyr isn't working. Most recently, it was because there
154# was a 64-bit RPM installed; remove it and install Joe's 32-bit one
[1205]155    YUM erase -y mit-zephyr
156    # mit-zephyr has a spurious dependency on mit-krb-config
157    yumdownloader mit-zephyr.i386
158    # if deps change, this breaks
159    YUM install -y libXaw.i586 libXext.i586 libXmu.i586 ncurses-libs.i586 readline.i58
160    rpm -i --nodeps mit-zephyr-2.1-6-linux.i386.rpm
[861]161
[1179]162# Install the athena-base, athena-lprng, and athena-lprng-misc RPMs
163# from the Athena 9 build (these are present in our yum repo).  Note
164# that you will have to use --nodeps for at least one of the lprng
165# ones because it thinks it needs the Athena hesiod RPM.  It doesn't
166# really.  Before doing this, run it without --nodeps and arrange to
167# install the rest of the things it really does depend on.  This will
168# include a bunch of 32-bit rpms; go ahead and install the .i586 versions
[1205]169# of them.
170    YUM install -y athena-base
171    YUM install -y athena-lprng
172    yumdownloader athena-lprng-misc
173    # ezyang: I couldn't find any deps for this that existed in the repos
174    # You might get a "find: `/usr/athena/info': No such file or directory"
175    # error; this is fine
176    rpm -i --nodeps athena-lprng-misc-9.4-0.i386.rpm
[1179]177
[1058]178# Install the full list of RPMs that users expect to be on the
[1205]179# scripts.mit.edu servers.
[387]180
[1205]181# ezyang: Running the below I got file conflicts. To fix (since I had
182# botched steps above), I manually compared package lists and installed
183# them.  If you've done the krb5 setup originally correctly, then
184# write down what you had to do here.
185    yumdownloader krb5-devel
186    rpm -i --force krb5-devel-1.6.3-20.fc11.i586.rpm
187    rpm -U --force krb5-devel-1.6.3-20.fc11.scripts.1138.x86_64.rpm
188    yumdownloader krb5-server
189    rpm -i --force krb5-server-1.6.3-20.fc11.scripts.1138.x86_64.rpm
190
191
192# on another server, run:
193rpm -qa --queryformat "%{Name}.%{Arch}\n" | sort > packages.txt
194# arrange for packages.txt to be passed to the server, then run:
195    # notice that yum is not capitalized
196    # Also notice skip-broken
197    NSS_NONLOCAL_IGNORE=1 cat packages.txt | xargs yum install -y --skip-broken
198
[1198]199# Check which packages are installed on your new server that are not
200# in the snapshot, and remove ones that aren't needed for some reason
201# on the new machine.  Otherwise, aside from bloat, you may end up
202# with undesirable things for security, like sendmail.
[1205]203    rpm -qa --queryformat "%{Name}.%{Arch}\n" | sort > newpackages.txt
204    diff -u packages.txt newpackages.txt  | less
205    # if all went well, you'll probably see multiple kernel versions
206    # as the only diff
207    # ezyang: I got exim installed as another package
[1198]208
[1058]209# Install the full list of perl modules that users expect to be on the
[1108]210# scripts.mit.edu servers.
[1058]211# - export PERL_MM_USE_DEFAULT=1
212# - Run 'cpan', accept the default configuration, and do 'o conf
213#   prerequisites_policy follow'.
214# - Parse the output of perldoc -u perllocal | grep head2 on an existing
215#   server, and "notest install" them from the cpan prompt.
[1109]216# TO DO THIS:
217# On another server, run:
[1179]218# perldoc -u perllocal | grep head2 | cut -f 3 -d '<' | cut -f 1 -d '|' | sort -u | perl -ne 'chomp; print "notest install $_\n" if system("rpm -q --whatprovides \"perl($_)\" >/dev/null 2>/dev/null")' > /mit/scripts/config/perl-packages.txt
[1109]219# Then on the server you're installing,
[1198]220#    cat perl-packages.txt | perl -MCPAN -e shell
[1206]221    export PERL_MM_USE_DEFAULT=1
222    # XXX: Some interactive gobbeldygook
223    cpan
224        o conf prerequisites_policy follow
225        o conf commit
226# on a reference server
227perldoc -u perllocal | grep head2 | cut -f 3 -d '<' | cut -f 1 -d '|' | sort -u | perl -ne 'chomp; print "notest install $_\n" if system("rpm -q --whatprovides \"perl($_)\" >/dev/null 2>/dev/null")' > perl-packages.txt
228# arrange for perl-packages.txt to be transferred to server
229    cat perl-packages.txt | perl -MCPAN -e shell
[812]230
[1058]231# Install the Python eggs and Ruby gems and PEAR/PECL doohickeys that are on
232# the other scripts.mit.edu servers and do not have RPMs.
[1206]233# The general mode of operation will be to run the "list" command
234# on both servers, see what the differences are, check if those diffs
235# are packaged up as rpms, and install them (rpm if possible, native otherwise)
[1179]236# - Look at /usr/lib/python2.6/site-packages and
237#           /usr/lib64/python2.6/site-packages for Python eggs and modules.
[1206]238#   There will be a lot of gunk that was installed from packages;
239#   easy-install.pth will tell you what was easy_installed.
[1179]240#   First use 'yum search' to see if the relevant package is now available
241#   as an RPM, and install that if it is.  If not, then use easy_install.
[1058]242# - Look at `gem list` for Ruby gems.
[1179]243#   Again, use 'yum search' and prefer RPMs, but failing that, 'gem install'.
[1206]244#       ezyang: rspec-rails depends on rspec, and will override the Yum
245#       package, so... don't use that RPM yet
[1058]246# - Look at `pear list` for Pear fruits (or whatever they're called).
[1179]247#   Yet again, 'yum search' for RPMs before resorting to 'pear install'.  Note
248#   that for things in the beta repo, you'll need 'pear install package-beta'.
[1206]249#   (you might get complaints about the php_scripts module; ignore them)
[1198]250# - Look at `pecl list` for PECL things.  'yum search', and if you must,
251#   'pecl install' needed items.
[1206]252    # Automating this... will require a lot of batonning between
253    # the servers. Probably best way to do it is to write an actual
254    # script.
[785]255
[1206]256# Setup some Python config
257    echo 'import site, os.path; site.addsitedir(os.path.expanduser("~/lib/python2.6/site-packages"))' > /usr/lib/python2.6/site-packages/00scripts-home.pth
[812]258
[1198]259# Build and install the scripts php module that enhances error logging info
260# XXX This thing really ought to be packaged
[1206]261    cp -r /srv/repository/server/common/oursrc/php_scripts /root
262    cd /root/php_scripts
263    ./build.sh
264    cp test/modules/scripts.so /usr/lib64/php/modules
[1198]265
[1179]266# Install the credentials.  There are a lot of things to remember here:
[1211]267#   o This will be different if you're setting up our build/update server.
[1179]268#   o You probably installed the machine keytab long ago
[1211]269    ls -l /etc/krb5.keytab
[1179]270#   o Use ktutil to combine the host/scripts.mit.edu and
271#     host/scripts-vhosts.mit.edu keys with host/this-server.mit.edu in
272#     the keytab.  Do not use 'k5srvutil change' on the combined keytab
[1211]273#     or you'll break the other servers. (real servers only)
[1179]274#   o The daemon.scripts keytab
[1211]275    ls -l /etc/daemon.keytab
276#   o The SSL cert private key (real servers only)
277#   o The LDAP password for the signup process (real servers only)
278#   o The SQL password for the signup process (real servers only)
279#   o The LDAP keytab for this server, which will be used later (real servers only)
280#   o Replace the ssh host keys with the ones common to all scripts servers (real servers only)
281#   o You'll install an LDAP certificate signed by the scripts CA later (real servers only)
[1179]282#   o Make sure root's .k5login is correct
[1211]283    cat /root/.k5login
284#   o Make sure logview's .k5login is correct (real servers only)
[387]285
[1058]286# If you are setting up a test server, pay attention to
287# /etc/sysconfig/network-scripts and do not bind scripts' IP address.
[1211]288# You will also need to modify:
289#   o /etc/ldap.conf
290#       add: host scripts.mit.edu
291#   o /etc/nss-ldapd.conf
292#       replace: uri *****
293#       with: uri ldap://scripts.mit.edu/
294#   o /etc/openldap/ldap.conf
295#       add: URI ldap://scripts.mit.edu/
296#            BASE dc=scripts,dc=mit,dc=edu
297#   o /etc/httpd/conf.d/vhost_ldap.conf
298#       replace: VhostLDAPUrl ****
299#       with: VhostLDAPUrl "ldap://18.181.0.46/ou=VirtualHosts,dc=scripts,dc=mit,dc=edu"
300# to use scripts.mit.edu instead of localhost.
301# XXX: someone should write sed scripts to do this
[562]302
[1211]303# If you are setting up a test server, afsagent's cronjob will attempt
304# to be renewing with the wrong credentials (daemon.scripts). Change this:
305    vim /home/afsagent/renew # replace all mentions of daemon.scripts.mit.edu
306
[1058]307# Install fedora-ds-base and set up replication (see ./HOWTO-SETUP-LDAP
308#   and ./fedora-ds-enable-ssl-and-kerberos.diff).
[785]309
[1058]310# Make the services dirsrv, nslcd, nscd, postfix, and httpd start at
311# boot. Run chkconfig to make sure the set of services to be run is
312# correct.
[1211]313    chkconfig dirsrv on
314    chkconfig nslcd on
315    chkconfig nscd on
316    chkconfig postfix on
317    chkconfig httpd on
[818]318
[1206]319# Postfix doesn't actually deliver mail; fix this
320    cd /etc/postfix
321    postmap virtual
[1179]322
[1058]323# Run fmtutil-sys --all, which does something that makes TeX work.
[1211]324    fmtutil-sys --all
325    # ezyang: I got errors on xetex
[803]326
[1058]327# Ensure that PHP isn't broken:
[1206]328    mkdir /tmp/sessions
329    chmod 01777 /tmp/sessions
[954]330
[1186]331# Ensure that fcgid isn't broken:
[1206]332    chmod 755 /var/run/httpd
333    chmod 755 /var/run/httpd/mod_fcgid
[1211]334    # ezyang: The latter didn't exist for me
[1186]335
[1211]336# Fix etc by making sure none of our config files got overwritten
337    cd /etc
338    svn status | grep M
339    # ezyang: I had to revert krb5.conf, nsswitch.conf and sysconfig/openafs
340
[1058]341# Reboot the machine to restore a consistent state, in case you
342# changed anything.
[1211]343    # ezyang: When I rebooted, the following things happened:
344    #   o Starting kdump failed (this is ok)
345    #   o postfix mailbombed us
346    #   o firstboot configuration screen popped up (ignored; manually will do
347    #     chkconfig after the fact)
[875]348
[1058]349# (Optional) Beat your head against a wall.
[562]350
[1058]351# Possibly perform other steps that I've neglected to put in this
352# document.
[1211]353#   o In the first install of not-backward, ThisCell got clobbered, resulting
354#     in trying to get tickets from openafs.org. Not sure when it got
355#     clobbered -- ezyang
356#   o For some reason, syslog-ng wasn't turning on automatically, so we weren't
357#     getting spew
358
359# Some info about changing hostnames: it appears to be in:
360#   o /etc/sysconfig/network
361#   o your lvm thingies; probably don't need to edit
Note: See TracBrowser for help on using the repository browser.