This is part of:

• Physics lab computers and status

TOC

See also

• Pertinent Incident Numbers:
  • INC000001186777 Physics Instruct - Met with Alan G. (Aug. 2, 2014)
  • INC000001709453 Physics Instruct - Image Testing computer hard drives "just in case". (June 30 - July 7, 2016)
  • INC000001883883 Physics Instruct - Testing Ctr increase robustness (Feb. 10, 2017)
    
    • Primarily to ensure Physics IT staff image the hard drives sooner than later.
• P123. A OneNote page started. Of any value beyond what we put here, in the wiki?

High-level recommendations

1) Recreate current system on 3 new computers with current operating system and versions of applications.

• Contemporary software ensures longer-term support.
• Contemporary operating system also ensures easier deployment on new hardware.
• Successfully recreating this interrelated system, and properly documenting it, ensures complete understanding of system, enabling correcting problems or improving the system more quickly and with less risk.

2) Clarify "system down" procedures for likely scenarios for current set-up (and then new system)

• Ensures appropriate response by Testing Center staff and IT support.
• Characterize "pain points" working around different types of failures to help calibrate levels of urgency.
  • Examples: What happens if failure persists for 1/2 day? A week? etc.
• Helps us all invest in activities we wished we did before a failure.
  • For example, Physics IT has made images of each of the 3 computers. What else might be done?

Recommended approach

The current system generally works well so serves as a good model, so focus on recreated it.

• We recognize there may be desired changes, nonetheless.

Oliver's recommended strategy is to recreate the current system's interrelated functionality using current OS and applications and use that learning to inform possible improvements. Can do this development before investing in new hardware. Deploy into production on new computers. New computers would cost about $650 with 4 year (or more?) warranties, so looking at about $2,000 god for about 4 (5?) years. Keep current scanners, printers, and networking.

Collect improvements desired from Testing Center  and IT Support staff.

Some improvements to consider:

• Improve FMPro scripts (if Oliver understood Nick's idea correctly).
• Have FMPro DB run on a dedicated system such that the two scanning systems are then identical to each other, making them even easier to support.

Determine who will lead the project

• Tracking work to be done and why, by whom, by when, who reviews/ vets work, etc.

Review what is known already.

Alan, Nick and other Testing Center staff clearly know a lot regarding not only the necessary desired results, who the technical functions of the system meet those objectives, and all related workflows. Physics IT has made efforts to characterize "what is" from a technical and work-flow perspective:

• Testing Center software relationships
• Testing Center workflow

Next steps

2017-02-10

• Nick and Oliver created some to-dos for each other. Completed ones appear in the "Status reports" section, below.
• Here are the outstanding to-do ideas:

Nick Taylor (Testing Lab) to dos:

Tech-related work (in priority order)

1. Determine how daily file-copies are being done from TC2 to TC3. In that case FMPro file copied. Determine how to replicate so copies also going to TC1. And copies from TC1 to TC2 and TC3.
   • All systems have access to "Backup" folder on each of the other systems (verified that today).
   • Consider all important files to copy over regularly, per:
     • Testing Center software relationships
     • Testing Center workflow
2. Determine how EZData functions, including dependent helper files, so that it could be reconstructed if on new OS with new version of FM Pro.
   • Current version of FM Pro used is v11.
   • Reads scanned test results and inputs them into FMPro, per:
     • Testing Center workflow
3. Determine how FMPro scripts work so system could be reconstructed or migrated to new version of FM Pro.
   • Current version of FM Pro used is v11.

Procedures and such

1. Establish clear procedures for staff to follow if a particular computer or printer is not working.
   1. How much to attempt to fix (answer: very little!). Who to contact (anser: <PhysIT@cornell.edu>). How to do things "by hand", and what to do after system is back up and running.
2. Question Nick was going to discuss with Alan: Is it worth have a weekly "snapshot" on a USB drive in Alan's office, which would be more "off site"? Worth the added risk?

Status reports

2017-03-09

Imaged systems now contained within an encrypted file, placed on a thumb drive, to give to Craig for safe-keeping.

Was: 2/10/17: Image all three systems in case of hard drive failure. Might protect against other likelihoods, too.

• KEY STEP: Carefully document where (storage device) with images are stored.
• Consider: Test restore of image to new hardware to see outcome. Will that work? (Alas, still Win7 not Win10. Still FM Pro v11, not v15.)

2017-03-01

Request from Craig for recommended action.

2017-02-23

Discovered that Brightworks is not available to do this work. (Was: 2/15/17, Oliver contacted Brightworks to see if they might be engaged with this work: Ticket#95300/Physics Testing Center, ensuring robustness in its setup.)

2017-02-10

Oliver and Nick Taylor <nwt2> created to-do lists following failing TC1 this week. Either in "Next steps" or "Status reports" (both, above), depending on completion level.

Past "Next steps"

8/27/15

• Completed 7/7/2016, all 2 systems, INC000001709453. Was:
  • To do: Image TC-01 and TC-02 to files. When shall we do this, this semester? Alan had discussed maybe removing grade info before doing this.
• Not completed:
  • To do: Get file synchronization to work. (What I said last time: I think I'm close to getting that to work across all 3 systems, so we'll see if I'm accurate on that count.)

Historical info:

Prior defined "next step", from spring 2015:

• Do this work in early summer 1015. Thus, reconnect in mid-May to coordinate this upgrade, using information on this site.

Sub-pages