Child pages
  • ramcloud.stanford.edu
Skip to end of metadata
Go to start of metadata

This page describes the configuration and uses of the ramcloud.stanford.edu server. It's a virtual server hosted by Linode. Collin Lee has access to the management console; several people have sudo access.

Web redirects and static pages

  • http server: redirects to https
  • https://ramcloud.stanford.edu redirects to RAMCloud wiki
  • /wiki/* redirects to RAMCloud Confluence wiki
  • /jira/ redirects to RAMCloud JIRA issue tracker
  • /~username/ serve static files found in /home/username/public_html/, e.g., https://ramcloud.stanford.edu/~ongaro/thesis.pdf
  • /~ongaro/raft.tla redirects to https://github.com/ongardie/raft.tla
  • /wiki/download/attachments/11370504/raft.pdf redirects to /raft.pdf (old location on old wiki, unfortunately indexed by google)
  • /raft.pdf redirects to /wiki/download/attachments/6586375/raft.pdf (current location on new wiki)

Dumpstr

this is a WSGI process started by apache

see https://github.com/ongardie/dumpstr/ for more info

IRC logs

nadoka records them using an IRC bot, then they are served by apache at /irclogs/.

started by a custom init script at /etc/init.d/ircbot

RAMCloud documentation

served at /docs/ by apache. generated by cron job under user "nightly".

Raft implementation table order

Cron job to fetch some information from github for use on raftconsensus.github.io.

see ~ongaro/raft/repos.jsonp,  ~/raftconsensus.github.io/fetch.sh

 

Previous uses

  • Confluence
  • JIRA
  • Postgres (for Confluence and Jira)
  • Raft user study (CGI pages for students to log in and take quizzes)
  • IRC lease bot
  • jungledisk server for backups (now rely on Linode backups instead)

 

Setup notes (Diego, Sep/Oct 2014)

removed exim4 and rpcbind (unnecessary open ports)
installed security updates
set hostname
set time zone
created user account for ongaro (with sudo access)
disabled password SSH login
installed and enabled unattended-upgrades
apt-get install build-essential git sqlite3 zip strace

apt-get install apache2
removed 000-default site
a2enmod ssl
a2enmod userdir
copied http to https redirect from old server
copied https config from old server. only enabled redirects to atlassian (rest TODO)
copied certs from old server

submitted helpsu ticket for updating DNS name (INC000003176758)
DNS updated around 2014-10-02 16:30 PDT
Requested reverse DNS update from Linode on 2014-10-03

Ryan moved his home directory over

apt-get install openjdk-7-jdk --no-install-recommends
apt-get install doxygen
apt-get install libprotobuf7 libprotobuf-dev protobuf-compiler
apt-get install libboost-dev
apt-get install libpcrecpp0 libpcre3-dev
apt-get install libzookeeper-mt-dev
apt-get install gcc-4.4 g++-4.4
apt-get install libboost-program-options-dev libboost-filesystem-dev libboost-system-dev
apt-get install libssl-dev
apt-get install python-epydoc
added nightly.sh job to crontab
enabled serving /docs/ over apache
note: docs won't update until DNS and dumpstr are moved over, since the script first tries to upload stats to dumpstr

created remaining user accounts
moved over user accounts (except dumpstr and ircbot)

shut down ircbot
copied ircbot home directory
copied /etc/init.d/ircbot
sudo update-rc.d ircbot defaults
sudo service ircbot start
updated apache config to serve /irclogs/

reverse DNS doesn't seem to work for IPv6 address
need a forward AAAA record first - requested and fulfilled 2014-10-14, then updated Linode

copied dumpstr home directory
sudo apt-get install libapache2-mod-wsgi
sudo apt-get install python-django
sudo apt-get install libjs-jquery libjs-jquery-fancybox libjs-flot libjs-jquery-cookie
updated apache config to serve /dumpstr/

apt-get install python-bs4
added raftconsensus cron job

created Linode account for Collin

 

  • No labels