Experience of installation of integration server for ATLAS DDM Dashboard

1. Machine dashb-ai-677 was created with the same parameters as dashb-ai-611 (existing integration one).
2. It was assigned to the same host group - dashboard/web_server/rucio/integration.
3. Consumer was started, but dirq2db service produced many errors on duplicated messages (previously inserted by service on dashb-ai-611). Consumer was stopped and Puppet parameter =dashboard_consumer_ was changed to "-" to prevent running the consumer temporarily.
4. Web interface is running successfuly, but no data is displayed.
5. All the configurations files seems to be the same as on existig dashb-ai-611.
6. Owner of /opt/dashboard/var is root:root. Should be be dboard:root, fixed manually (TODO: configure with Puppet). After the fix web interface is working properly.
7. Database to be used: ddmdevel@INT11R. Following scripts were ran from DDM DAO to initialize it: ddm-oracle-objects.sql, ddm-oracle-packages.sql, ddm-oracle-admin-packages.sql. Other are not needed for the development database (e.g. privileges, synonyms and alarm jobs).
8. To prevent Puppet overwriting dashboard-dao.cfg, dashboard_dao parameter was set to "#".
9. For oracle and dao_ddm_w section credentials were set to the ddmdevel ones.
10. It took ~2 hours to have tables initially filled to have plots on the Web UI (despite dirq2db collector was saying that events are being inserted; it seems that the database was slow that time).
11. File /opt/dashboard/cron/dashbConsumers.config is not created automatically. Created manually. Issue was put to JIRA.

-- SergeyBelov - 2015-03-11