Standup call notes 2013 Block 2-3 ========================== Skye * 20130412 * Noticed marshalling exceptions in dev due to null contact subjects - for Rob * Replication auditing meeeting * Generating reports for bad replicas on MNs and CNs * Fixes to the stage indexing noticed from SEAD * 20130410 * Replication auditing meeting * need to discuss delete() or archive() usage * Adding in activity logging into auditing * Troubleshooting SEAD indexing * ORE processing is asynchronous * indexing tries to make sure referenced objects are present * big OREs are troublesome, but Solr URLs can be chunked * however, we allow packages of unbounded size * meltdown imminent * need to discuss hierarchical organization of packages * 20130408 * Planning on meeting for dealing with invalid replicas * In touch with Kavitha (SEAD) and Ryan (Dryad) on a DSpace stack * 20130405 * #3514 - installed in dev now * Assigned a metacat patch to matt - #3699 * Need a meeting on auditing for dealing with CN bad replicas. will send out an email * DSpace is installed, running locally * Mutability will be a big factor * Saving past versions isn't currently supported * 20130403 * Auditing testing in dev. * Found bug in CNodeService preventing audit from updating verification date. * TODO: send patch to Matt * Found invalid original copies of objects, interesting - on demo5. same pids/different checksums * DSpace - coming up to speed on it before planning D1 MN implementation * Redmine 3514 * 20130401 * Oddities in replication being investigated * side effect of not using failure factor * saw serial version of in system metadata being increased when no replication occured * Auditing testing, going to test the whole process in dev * Started looking at DSpace tasks * Need to get the story entered into Redmine and get tasks de-composed. * What parts can we re-use from Dryad, SEAD DSpace implementations Roger * 20130412 * Continued ONEDrive mockups * 20130410 * Working on mockups for ONEDrive * Using Fuse - sets up a directory structure, using screenshots * Two basic types of functionality in mockups * object discovery * presentation of objects * Usability group is meeting Apr/May shoulder * 20130408 * Finished up suggested changes to MN checklist * fairly extensive changes * simplified upper levels * added details into linked documents * Re-wrote text, worked on logical order * 20130405 * Working on MN deployment guide * Will be creating mockups for ONEDrive soon * 20130403 * checking MN deployment checklist wrt GMN notes * has many edits and restructures - * TODO: commit as a subdirectory of mn deployment directory * ONEdrive - finished phase 1 is complete, but not user friendly * looking for input from Usability Group before moving forward * plan to meet with them at the end of the month * Dave's interface shown to scientists is probably close, but rough. * multiple browse hierarchies * what to do at the leaf nodes - how to present * discovery through the interface * TODO: mockups for next week to discuss * 20130401 * Start to focus on redmine story for ONEDrive work breakdown * Will discuss with Chris B. Rob * 20130412 * Finished mutability proposal document * Working with Soren/EDAC - trouble with ORE format * Will work on server side API calls for version history access * Ryan will work on Dryad node next * 20130410 * writing up the mutability proposal * 20130408 * Updating d1_integrations tests * Working with Soren on syncFailed() and getReplica() calls * They aren't currently using ORE serialization for resource maps, will work with them on this * Working on mutability document * Dryad: had sent some bug fix requests to Ryan * some character encodings are incorrect for getLog() * he'll get to them after they deploy their new web UI * how are they going to deal with old versions? * 20130405 * Webtester work - getReplica() test * Uses a public doc, isNodeAuthorized() isn't called, so the env dependency isn't as complex in testing * Working with Soren - EDAC * need a test certificate * synchronizationFailed() implementation - is throwing not implemented, but should be implemented * can do an IP check on it to stop spam if they're not using https + certs * identifiers - do we recommend that they use reserveIdentifier()? * 20130401 * Working with Soren on EDAC * How they are formatting exceptions and response types * Discussed mutability and series ID with Dave * Incorporating in mutability epad: http://epad.dataone.org/201301-mutability * Added a redmine feature: 3698 * Issues with what is returned from different nodes can vary based on where they are on the sync and replication cycles Chris B. * 20130412 * ONEDrive and Ansible development: waiting on schema changes from Robert * Documentation work for when Chris leaves in August * 20130408 * Network outage today - 10pm * DNS entry for splunk-orc.test.dataone.org * 20130408 * Chris Jones and I will be handling outage at UT. * Scheduled for 4/10 10pm EDT to the morning of 4/11 * Stopping the dataone services and taking the cn out of round robin. * Operations email waiting approval * Splunk * Feeding syslog data of cn-dev-orc-1 to splunk. * Need a DNS entry and a cert for the splunk server. * Ansible development stuck * Waiting on schema from Robert. * 20130403 * OIT has rescheduled the network shutdown for April 10th * impacts everything at ORC - all VM's unaccessible for several hours. * UT campus, not Oak Ridge campus * TODO: get timing info, and send notification to operations mailing list * find the plan that was going to be used in the fall * 20130401 * Put Splunk behind SSL * Need to see about obtaining an "official" SSL cert. * Completed move to Knoxville (no more going back to Nashville) * Waiting on Robert Waltz for XML schema to finish Ansible playbooks * Need to Sync with Roger on OneDrive work Ben L. * 20130412 * Morpho update: replace "Owner" wording with "Creator" to prevent negative connotation of "data ownership" and also to align with EML field name that stores it. * TFRI approved in cn-stage, but there is an SSL peer unverified exception (surprise surprise). I believe it has to do with them using a DataONE Test CA issued server certificate. Have instructed Meei-ru to include the CA chain file in their Apache directives. * Morpho/Kepler (TFRI commiunication): * I've instructed Chau Chin Lin how to change the language used in both Morpho and Kepler (uses the OS/system locale). He wants the ability to run those programs in English even if the rest of his computer is running Chinese. I believe his team even modified their version of Morpho to allow this. It is certainly possible to set the locale for Morpho on start-up to a pre-configured locale. Perhaps add this to the advanced configuration screen. * cn-dev environment upgraded to use Metacat 2.0.6 on all three CNs. Awaiting test results before tagging release from the 2.0.6 branch. * Will be meeting with semantics WG next week as they are in town at NCEAS (mon-wed) * 20130410 * TFRI MN is back, working with Meei-ru * Will work in stage * Unicode: language issues were with Morpho/Kepler and the translations we have, not any technological issue. * 20130408 * Identity Service and /portal issues WRT UTF-7, escaping, LDAP storage, etc. * working on cn-dev * 20130405 * Working on releasing Metacat 2.0.6 * Working with TFRI * DataUp meeting in SB - soliciting ideas on writing a proposal to further the work * 20130403 not present * 20130401 * Replication Policy in Morpho wizard * Preparing for metacat release, might be issue with deployment on CN that needs to be further tested (regarding updateSystemMetadata() not always being called) Matt * 20130410 * Reviewing the MN Checklist document * Helping Kavitha with SEAD questions * Mostly focusing on other projects * Plone sites going down every 24 hours or so * Nick's working on it * 20130408 * MN checklist work needs consolidation * concensus is to keep it in subversion for now * Laura will focus on Coordination sections * some feedback - still confusing * 20130403 nothing to report * 20130401 * Checklist for MN Deployment coming along, still needs some details, but its probably useful now * http://mule1.dataone.org/OperationDocs/member_node_deployment/mn_checklist.html * http://ask.dataone.org is going live Robert * 20130412 * Working on CN buildout scripts, will meet with Chris B today * CNs are 8 pids out of sync on cn-orc-1 - will try to resync * Shibboleth will be deployed at UT in May (rumor?) * 20130405 * DataUp meeting in SB * 20130403 not present * 20130401 not present Chris J. * 20130412 * Reviewing mutability proposal * Working on broadcast of system metadata change in CNodeService in Metacat (dev env) * SEAD work with Kavitha * MN Wranglers discussion with Laura, John, Rebecca, etc. * 20130410 * working on the SEAD MN deployment in stage * CN-CN replication issues: look at Metacat timed replication config * formatid issues * count=0 bug * 217 of 218 objects indexed so far, but indices are out of sync * ORC not indexing * TODO: DNS entry for splunk-orc.test.dataone.org * 20130408 * troubleshooting metacat CNodeService changes * Working with Kavitha on SEAD * TODO: change formats to only allow OREs as RESOURCE * TODO: EDAC test certs * TODO: Dryad ticket for checksum() uniqueness * 20130403 not present * 20130401 - (not present) (in ABQ for Sensor Net meeting) Dave V. * 20130412 * Looking into RSS feeds in the sphinx documentaion * Mutability doc review * Semantics * Documentation: allocate a week to bring it up to date - week after next * also need to create versioned documentation * 20130410 * Backlog * Worked on the DataPackaging documentation * 20130408 * Writing a proposal for DataUp * Working through ticket backlog * TODO: fix documentation to only support ORE serialization for now * 20130405 * Just getting back in * Usability of ONEDrive * Questions on ONEDrive use of foresite? * Roger - finished that work up * 20130403 not present * 20130401 not present Discussion on MN checklist * Diagram suggested changes: generate via Sphinx and graphviz * Dave: Powerpoint is easier for other users, should stick with it * Selecting a Tier should be part of planning * Factored out common pages (like registration) and pointed to it from multiple higher pages (prod registration, test registration) * * Simplifying the "Development" or "Build" section * Plan * Test? * Deploy * Operate XHDY6THAKK