This is an old revision of the document!

---

Kumu is an on-boarding customer.

We have 1 lab relay and 3 production relays.

There is an AWS instance that is running our production trial environment.

• kumuapps1.err
• kumuapps2.err
• kumuesk1.err
• kumulb.err
• kumulbdbrmq.err
• kumuoat1.err
• kumuoat2.err

Currently, the system is only running the MDC agent on kumuoat1.err. kumuoat2.err functionality has been shut down (more on that later).

The MDC purely runs on kumuoat1.err, though we have a strange VPN currently. If the kumuoat1.err server is restarted, please stop all MDC services:

1. sudo systemctl stop orchestrator;
2. sudo systemctl stop tracker;
3. sudo pkill -f rdfagent;

Then run:

1. sudo semanage permissive -a ipsec_t
2. sudo sysctl -w net.ipv4.tcp_timestamps=0
3. sudo systemctl start strongswan

Make sure you get positive ping results from:

1. ping 10.8.1.8

or the like, some production IP, which is available in the MySQL orch db at kumuesk1.err element table.

Contact shea.lawrence@errigal.com (415-559-8389) if you have any questions.

NOTE: If disk space fills up, then you need to delete the /var/log/messages file most likely, and may have to restart the instance through AWS. The instance is in the 'aws_sandbox' login in US East Ohio. If you need access, contact Shea Lawrence or Stephen Shotts.

NOTE: Currently, please do not start up the MDC on kumuoat2. The playbook may do that. We are simply in a period where running on both may cause complications, which is being investigated. Thanks!