Tasks for Implementing Unidata Internet Data Distribution (IDD)

Related Information:

• What's New with the IDD.
• Main Unidata IDD Page.

Very High Priority Indeed (VHPI) Tasks

• Contacted each site not participating in the IDD via telephone or e-mail to notify the site representative of the cessation of the Unidata/Wisconsin satellite broadcast Dec 31, 1995.
• Contact remaining sites that are not connected to the IDD. Determine status and attempt to facilitate deployment. Check status for latest deployment information. (Linda M - in process).
• Get Solaris running on source LDM machines (Alden).
• Renew efforts to get alternate FOS source site running: either get CSU running or try to get a group of Ku subscribers to take responsibility for individual feeds (Ben).
• Develop better troubleshooting and monitoring systems to help narrow down the cause of deliver problems (Mitch and Robb, Netcheck is running at most relay nodes; Syscheck is being deployed; system for simplifying LDM log output is on hold because it consumes to many system resources).
• Track down cause of infrequent problems with failed ingesters and corrupted queues on Alden machines.(Glenn and Steve with Mitch and Robb, DONE.)
• Work with Alden to get all OS patches into their IDD machines. (Mitch and Robb with Karen Wallace. DONE.)
• Start major effort to inform the community that is not using the IDD that they have to get onto the system before the end of the year (Newsletter cover article addresses this. Work will be ongoing throughout year.)
• Fix colors and labeling for IDD reliability charts. (Robb) (DONE)
• Develop graphical displays of hourly and daily reliability statistics (Robb). (DONE, 12/30/94)
• Redesign routing topology now that we have experience with a number of relay nodes (Mitch).
• After the new routing is in place, begin bringing online sites who delayed implementation for one reason or another (Mitch).
• Experiment with routing SUNYA data through Cornell. (DONE, 11/17.)
• Reroute New York State area traffic through Cornell at least until the networking into the SUNY sites is improved. (Mitch) (DONE)
• Find out what's happened to sites which reported statistics for a time and then disappeared from the statistics logs (Linda M.). (DONE, 12/30/94)
• Establish an alternative source for FOS to take over if the network out of Alden fails (Ben with Steve Finley at CSU as of Nov 2).
• Get reliable, well-connected relay node in Northeast (Ben working with BBN, Mitch working with Cornell). (Mitch's part is DONE.)
• Get remaining top-level relays running (Mitch, Robb). (LSU now running along with Cornell, Georgia Tech has problem with impenetrable firewall)
• Test and prepare final release of Mctingest for OS2 Sites (Tom, Steve). (Mctingest DONE and running for sites with FTP, Inc. networking; McLDM DONE and running at sites with IBM networking for OS/2)
• Prepare and test general release by Nov 1. (Glenn, Mitch, DONE 10/31).
• Work with Texas A&M to get around firewall problems (Mitch, DONE 11/2).
• Prepare final version of LDM Documentation by Nov. 1 (Matt, DONE 11/2).
• Get out announcements to community re deployment (Mitch, Ben, Dave, DONE 11/1).
• Rework system for determining working topology from the LDM logs at sites (Robb). (DONE)
• Contact the sites in the deployment queues we haven't heard from yet (Linda M.). (DONE, see IDD Site Deployment Status.)

High Priority Tasks

• Follow up on initiative to get IDD sites working together as regional teams to address the issue of system administration load (Ben).
• Get remaining (postponed) sites online. (Mitch, ongoing effort, priority increased, 5/95)
• Fix problems with statiics reporting on daily charts (Robb, DONE 4/95)
• Provide documentation for the IDD reliability charts and add charts that include only the relay node. (Robb, DONE 4/95)
• Fix LDM documentation to recommend correct time for mailing in statistics reports (Steve). (DONE)
• Prepare documentation on data recovery sites (Matt).
• Contact Bill Highlands about 24-hour montoring at Alden (Ben). (DONE)
• Get WARNING and MONITORING systems into source sites (Robb, nearly complete, 5/95)).
• Get additional relay nodes running, including mctingest sites (Mitch, Tom). (DONE)
• Investigate why so many sites, including all the new deployment sites, are not reporting statistics. (Robb, Mitch) (DONE)
• Investigate data delivery problems that cropped up at Miami and Naval Postgraduate School the weekend of 1/5/94. (DONE, both sites running reliably)
• Work with MIT on alternative to Ultrix (Mitch, Ben).
• Work with Rutgers on alternative to LINUX (Glenn, Ben). (DONE, found a Sun for relay node).
• Work with NC State on SunOS upgrade (Chiz). (DONE)
• Alabama Huntsville needs to change to Merit/Michigan for its upstream node. NC State will be taking their IDD system down for an OS upgrade at some point. (DONE)
• Make sure all test sites (except those with Ultrix and LINUX systems) convert to LDM 4.1.
• We have a rather comprehensive list of regional network providers. But now we need to get a list of who we contact in order to report a suspected problem in each region (Linda M.).
• Sites which need to upgrade to LDM 4.1:
  • Oklahoma/NSSL
  • Alabama Huntsville (also has to switch upstream feed from NC State to Merit) (DONE)
  • SUNY Brockport (UGRADED TO 4.1, 11/3)
  • Delaware
  • Western Michigan
  • Northern Illinois (UGRADED TO 4.1, 11/1)
  • Arizona (Ultrix problem) (DONE, new machine)
  • San Francisco State
  • New Mexico Tech (UPGRADED TO 4.1 11/9)
  • NSF/ATM (UGRADED TO 4.1, 11/1)
  • Rhode Island (?)
  • MIT (Ultrix problem)

Medium Priority Tasks

• Establish and document procedures to enable site administrators to troubleshoot problems, to check out underlying Internet connections, and to take appropriate actions to correct problems (Mitch, Robb, Ben, priority increased 4/95).
• Investigate statistics anomalies where relay nodes reported receiving less than 100% of products, but downstream nodes reported 100% reception (Robb, DONE and problem corrected, 4/95).
• Develop OS/2 LDM ingest module to obviate need for Mctingest (Tom, Steve). (DONE)
• Work with ANS and SUNYA to get Internet upgrade in NY region (Ben with Mark Oros of NYSERNET and Guy Almes of ANS). (ROUTED AROUND PSI BOTTLENECK).
• Work with ANS to monitor and chart changes in Internet topology (Ben, Mitch, Linda M.).
• Continue effort to get all sites onto Internet (Linda M.).
• Enhance hourly and daily statistics reports to include sites for which stats are not available (Robb).
• Investigate statistics reporting problems at Naval Postgraduate School (Robb). (FIXED)
• Investigate redundant product transfers (showing up as hourly delivery of more thatn 100% of products) at Florida State. (FIXED, 10/31) (DONE)
• Fix filter to remove partial site list at the beginning of the hourly IDD statistics summary (FIXED, 11/17). (DONE)
• Sites with statistics reporting problems:
  • Parks College (FIXED)
  • University of Nevada Reno/DRI (FIXED 11/3)
  • University of Utah
  • Iowa State (FIXED)
  • Oklahoma/NSSL (Stats disappeared)
  • SUNY Brockport (FIXED, 11/2)
  • Western Michigan
  • Texas A&M (FIXED: 10/31)
  • Georgia Tech (Firewall problem).
  • Purdue (Stats come and go). (FIXED)
  • Yale (FIREWALL PROBLEM FIXED).
  • LSU (FIXED).
  • New Mexico Tech (May have network problems).
  • Auburn (May have network problems).
  • Naval Postgraduate School (strange name and numbers, NUMBERS PROBLEM FIXED, 11/2, name still needs work ALL FIXED, 11/21.)
  • Arizona (FIXED)
  • Michigan/Merit (FIXED: 10/30)
  • MIT (anomalous numbers, FIXED: 10/31)
  • Florida State (anomalous numbers, FIXED: 10/31)

Low Priority Tasks

Very Low Priority Tasks

• Enhance and port Mctingest to other platforms