...

1. Report from the Mozilla Global Sprint
   1. Learned a lot from the open leaders program. Will pull together the resources from that experience for other projects and for DuraCloud to pay attention to being open and inviting and having resources to share. Global sprint was a little disappointing, so one dropped in but it was worth the planning. We learned we need to be really intentional about engagement and reaching out. Not terribly surprising but useful to know. Could get a lot done in a small period of focused time on a DuraCloud sprint. Background work for the sprint was valuable. Person hours used? Could look at hours to see. Could be done more quickly now with template in place. Timing was poor and outside our control, so one planned with this group would be much better.
2. Communicating as a group (email, slack, gitter, etc)
   1. Used Gitter for the sprint, but within DuraSpace Slack is more commonly used and would be easier. Is that true for this group? Everyone is in Slack already, lowest barrier. Separate Slack channel, or communicate via email? Email for now while the group is small, and could invite to a Slack channel.
3. Working together
   
   1. Sprints or more independent work? How to coordinate and collaborate? Trying to understand our varying priorities and what could move us forward. Could do a development sprint or a higher level sprint that would give an opportunity to do some prioritizing and organizing. Having user feedback ahead of a sprint would be useful for prioritizing. 1 or two day sprint for prioritizing needs and a coordinated development sprint, but it is all a matter of timing. Development deep dives would be useful, a two week sprint or a little longer to figure out how long things take. 1 day planning dash to share user feedback and discuss priorities, and then follow with a development sprint looking at priorities from the dash. 
   2. October 5th for planning dash, October 15-26 for development work
4. Development priorities
   1. DuraSpace working on transitioning to multipart upload on file transfers. Partway into that work, hoping to be done by fall. Also sprint in place for updating streaming from Flash based to HTTP livestreaming. 
   2. Customers have been asking for search feature. What does that mean and would that be a priority? Tag and file name search. File name search would be really helpful for TDL as well. Might be better framed as a find feature. 
   3. Other features that are requested? Duplication among regions? 4Science is looking at this. Question if this is the best approach, as Amazon built-in application for cross-region replication might be a better approach. Bill: would want to do some verification. 4Science will update on this. 
   4. Operational needs: TDL: more informative error handling, better sync tool tooltips. Release notes for new versions, with a list of changes. Options to customize emails for registration in management console. Nick will add to JIRA.
5. Secondary storage providers: 4Science working on setting up support for another European provider, which also allows multipart upload. Andrea will provide an update on what they support. New kind of S3 storage that is not replicated has been introduced, is only in one data center and is cheaper, could be a good option to pair with another provider location. 4Science is looking into this. It is 25% less expensive than standard S3. 
   1. Some TDL customers want to be able to create spaces and push content to secondary providers. Issue comes up often with one client. Currently it is set up for replication only to simplify the workflow. REST API can be used to create spaces on the secondary provider if they are an admin. Taken out of UI. Bit checking is not being done on Glacier, so that is why there is not a Glacier only option. TDL does have customers with Glacier as a primary and it has caused issues. 
   2. Current way that we limit secondary providers isn't ideal. Something to discuss on the next call, how users want to engage with secondary providers. DuraSpace has heard questions from prospects but not from current customers; current customers are focused on secondary storage as a backup. Using Glacier in other regions and as a secondary store.
   3. Providers worth considering, example: Oracle supports full API but have not done any testing.
6. TDL and 4Science experience with use and implementation of DuraCloud: important to discuss on our next call.

Actions

•  Nicholas Woodward to add TDL operational needs to JIRA (more informative error handling, better sync tool tooltips, options to customize emails for registration in management console, etc)