Thalia Operations Plan notes July 19, 2007

Operational plans

monitoring, escalation procedures...

runbook in wiki - what ops needs to do when something happens

list of machines, where they are located, what they are used for

zinc operationsal wiki space.

ping service

Tomcat plugin for Nagios for ops?

admins need to be auditable.

status of cluster

Build processes need to be more resilient

jira - builds need to be neutrl wrt machine - versionoing should not depend on who does the build...

Known issues

e-mail with location of meeting notes.

tag search may require a new implementation of tags

If we delay going out to customers, we delay feedback for 1.0.

Give people a url for help

clean up url

Operations plan:

runbook by Qing and Hunter
link to runbook from wiki space
contact voip people for phone number that sends voicemail to thalia-dev. - Arnis
need understanding with ops on what monitoring gets done.
- server is up, apache is up, tomcat is up (provide module), alfresco is up (provide url), thalia (give them a url to look for), mySQL? - Hunter/Andrew
need ops to know who to notify: email thalia-ops@mit.edu
if it is down, restart service, if it stays down reboot. If it stays down, call us

Support plan:

from the app, people should have a link to report a bug
they need a phone number of e-mail address or access RT queue
Get a pone number that forwards to team members.
thalia-support does not have e-mail - put lists under isda-list-admin
thalia-support for reporting bugs
thalia-request for requesting domains
thalia-copyrights for report copyright infringements
terms of use
- MIT not responsible for content, individuals are responsible for content posted to the service. If we receive reports of copyright infringement,

stopit@mit.edu - copyright question. link to mit policy.

need chron job for backups - jira bug?

User acceptance extended