People often ask us how we handle tickets. Who answers them? What kind of support do we provide? How do we interact with other support services? Here are the answers to those questions.
When you send email to email firstname.lastname@example.org, two things happens. First, that mailing list is monitored by several people associated with the VDT. Second, it is forwarded to our ticket-handling system (called RT) at email@example.com. You can browse these tickets.
Once the ticket has been created, the core VDT team looks at it, and anyone on the VDT team is welcome to assign the ticket to themselves. The technical head of the VDT team (Alain Roy) takes responsibility to ensure that the ticket is assigned to someone in a timely fashion.
Our answers to the tickets are on a best-effort basis. We work regular business hours, so people should not assume that we provide help twenty-four hours a day, seven days a week--we simply do not have enough staff to provide that kind of service.
Sometimes we get lots of requests for support. When we do, we have to prioritize the tickets. People directly associated with our funding sources will get higher priority service than other VDT users. If you send us email and if there is a chance that we won't recognize you as a person associated with our funding, please let us know. That said, we do try to give everyone support.
The VDT groups sometimes gets questions that it can't answer. We don't know everything after all. When this happens, we try to find someone else to help us out. If it's a Grid3 or OSG infrastructure problem, we send a request for help to the iGOC. If we find a serious problem with Globus, we submit a ticket to the Globus Bugzilla. If we find a problem with Condor, we interact directly with the Condor project since the core VDT group is also part of Condor. For other pieces of software, we tend to email the author of the software directly. We will copy all relevant email to the person or organization that reported the problem--we will do our best not be a bottleneck, but will hook people directly together while monitoring the communication and helping out where we can.
When the iGOC sends us a request for help, we respond both to the iGOC and to the user that initiated the request to the iGOC, so that we can keep everyone in the loop.
Summary: When we get a request for help:
Interaction with OSG is a subset of what is desribed above--OSG is not a special case. Here is an example interaction that occured with an OSG ticket from the iGOC.