Skip to content

Latest commit

 

History

History
60 lines (39 loc) · 1.58 KB

data.rst

File metadata and controls

60 lines (39 loc) · 1.58 KB

Data storage and retrieval

Most sites have fairly simple data-layers. The notable exception is Socorro.

Typically we use some form of mysql with master-slave replication.

For search either Sphinx or Elastic Search are used.

For cache memcache and redis.

Socorro uses postgres and HBase.

Production Data

Sometimes having production-like data is necessary for debugging. Private data relating to users must remain on the Mozilla network. Therefore, there are two options for getting Production or production-like MySQL data.

Anonymous Data

In ~ddash/anonymize On webdev1.db.scl3.mozilla.com on the Mozilla/MPT VPN are anonymized dumps of production data for:

  • AMO
  • FlightDeck
  • SUMO

While the datasets are anonymous, they are not for general distribution.

Input Data

You can get a copy of the Firefox Input database by using the following script:

set -e
FILE=input_mozilla_com.`date +%Y.%m.%d`.sql.gz
USERNAME=username
LOCAL_DB=firefox_input

scp $USERNAME@webdev1.db.scl3.mozilla.com:~ddash/input_mozilla_com/$FILE .
zgrep -v "INSERT INTO \`feedback_term\`" $FILE > tmp.sql
cat tmp.sql |mysql -u root $LOCAL_DB
rm tmp.sql $FILE

Be sure to replace username with your actual LDAP username.

Webdev Database Cluster

Alternately, many production databases have copies running on webdev1.db.scl3.mozilla.com and webdev2.db.scl3.mozilla.com. You can connect directly to these servers.