QA Home Page

General Goals

• Monitor Apache for error 500, target 0.2% a month.
• No more than 2 data repushes a month.
• No more than 1 hour downtime during normal business hours.

Genome Browser Setup

• See Jorge's Presentation on the server setup.
• External site about our architecture: http://oliverelliott.org/article/bioinformatics/tut_genomebrowser/

GIT (Source Code Management System)

• Getting_Started_With_Git
• Working_with_Git
• Once you've read up...ask for git commit access.

Editing Static Docs

• Editing static pages that are destined for the RR (Round Robin), hgdownload, or linked to genomewiki.
• Need to have links hard coded to the RR, but need to test them on the development machines? Take a look at how you can use javaScript with the static pages.

CGI Build Process & Testing

• Currently we are on a 3 week CGI software release schedule. See CGI_Build_Schedule for more details.
  • The google calendar version should be added to your own google calendar. More information.
• The buildmeister follows this process.
• When the CGIs are loaded on hgwbeta, the QA'er tests their assigned CGIs. The general CGI testing instructions can be found here.
• Currently we try to test our CGI's in Internet Explorer & Firefox due to the high browser usage.
  • Brand new CGIs should be tested on the latest version of IE and FF.
  • All CGIs should be tested using the latest version of IE after they are built on hgwbeta.
  • On release day QA'ers test using their browser of choice, except the designated IE tester.
• If you don't have access to a Windows machine with Internet Explorer, there is a machine you can log into.

Cell Browser release process

Instructions for releasing the Cell Browser are found Cell Browser Release Process

GBiB and GBiC release process

To see how the GBiB and GBiC files are tested see more at the CGI_Testing page.

QAing tracks/assemblies

• Browser QA process:
  • New Assemblies
  • New track checklist
  • Specific Genome Browser Tracks

• Encode QA process

Mailing List Maven

• Mailing list schedule can be found on the Genome Browser Mail List soe google calendar. More information.
• Mailing list procedure can be found on the redmine wiki
• Places to search for answers:
  • Mailing list archives (which can be searched from these two pages on the main website: Contact and FAQ
  • Search the mailing list forum which isn't as extensive as the mailing list archives.
  • Public Genome Wiki
  • Search the redmine mailing list project for internal chatter on previously answered mailing list questions.
  • Frequently_asked_mailing_list_questions
  • List of Subject Matter Experts to go to for help.
• Remember that when you email users, make sure the links don't have session id's.
  • Ex: w/ session id: http://genome.ucsc.edu/cgi-bin/hgTrackUi?hgsid=190182411&c=chrX&g=geneid (remove the bolded section)
  • Ex: w/o session id: http://genome.ucsc.edu/cgi-bin/hgTrackUi?db=hg19&g=geneid (note that links to description pages should include the database: db=hg19&)
• Most other links without an hgsid will likely not work. Instead, create a session.
• If you click on a link supplied by a user that includes an hgsid, do not change any settings until you reset your cart.
• Make sure none of the links you respond with are to our internal servers, such as hgwdev. Typically we should only send links to genome.ucsc.edu, genome-preview.ucsc.edu, and to hgdownload.soe.ucsc.edu.

Updating Blat Servers

See this page for more information.

Tools

• Vi
  • Don't have a text editor of choice? Learn how to use vi by typing in "vimtutor" on the command line on hgwdev.
  • Vi Quick Start Guide
  • Handy Cheat Sheet

• UNIX Commands
• MySQL Commands
• Common Scripts
• Wiki Cheat Sheet