War Games for Sysadmins

The beginning of a calendar year is a great time to make wild predictions, so here’s one…within the next 10 years we’ll see a new class of system administration tools that allow freelance sysadmins to virtually team with an organization’s existing IT staff to help debug operational problems within that organization’s applications and/or infrastructure.

Key features I’d expect to see in these tools:

  • Ability to take a snapshot of an organization’s live environment, or a portion of it, and run that snapshot in the cloud. The environment would have all PII and other identifying info (e.g. IP addresses) removed to ensure that the sysadmins would not know on whose infrastructure they were operating. The result is that the freelancers would have their own sandbox in which to experiment, debug, and come up with potential solutions for a stated problem.
  • Ability to set parameters on which freelancers may work on a given type of problem, including allowing only pre-qualified freelancers (e.g. Resources from consulting firms with whom the organization already has a relationship)
  • Ability for freelancers to ask questions of the requesting organization about symptoms of the problem, or other background information that might be helpful in the debugging process.
  • Ability to screen share the snapshot environment and ability to pass control to the organization’s sysadmins, or to a member of a freelancer team.
  • Teaming capability where multiple freelancers can pool their talents to try and solve a particularly multi-disciplinary problem(e.g. A sysadmin and a network guy team up to debug a thorny NFS issue).
  • Replay capability so that the organization can see the solution implemented by the freelancers and validate that it would indeed meet their needs before compensating the freelancer (and before implementing the changes in their production environment!).
  • Identify common solutions to common problems and institutionalize that in the system to reduce time to resolution of those problems. Allow a requesting organization to potentially solve common issues themselves without having to involve the freelancers.
  • Payment system to ensure that freelancers are compensated if the debugging steps they suggested are successful.
  • Rating system for freelancers and participating organizations to ensure a high quality, trustworthy marketplace.

I don’t think that corporations would be willing to put their entire infrastructure stack out to be examined by a bunch of nameless, faceless freelancers. However, I think they would be willing to take specific applications with well understood datasets, like a problematic MS Exchange server, for example, and allow a PII/anonymized version of it to be worked on by a set of vetted, trusted freelancers. As I’ve said before, system administration is becoming more social, and over time it’s not unreasonable to think that ‘social’ would extend to on demand outsourcing of infra and app debugging with a virtual team component.


Leave a Reply

You May Also Like