Dear Small-Scale Admins,

Thank you for the advice during last week's meeting.

In the event we choose Galaxy (which we are leaning towards), I would be responsible for setting up the Galaxy instance as well as the initial maintenance.
While a sysadmin would do the long-term maintenance.
As such, we were wondering, in your experience, for those who wish to answer:
  1. How many hours per week are necessary for purely admin-related maintenance on a local Galaxy instance (used by let's say 50 users) ?
    • Not the time for developing custom tools, as users would be expected to do that (they already do this outside of Galaxy)
    • Not the time for end-user assistance, as we plan to create in-house tutorials
    • Just the time necessary for e.g. restarting Galaxy, creating a back-up of it's database, repairing in case of a crash, ...
    • Ideally this time would be low or even zero outside of upgrades. But that is probably being idealistic.
  2. How stable is Galaxy? I imagine crashes are rare.
  3. Are there any time-consuming admin. tasks that I am not aware of/did not mention above?
  4. Galaxy version upgrades
    1. How often is it recommended to upgrade the Galaxy version? 
    2. What are the consequences of not upgrading for a while, apart from Python min/max version incompatibility?
    3. How much time does an upgrade usually take?
    4. I am also worried that if I fork Galaxy to modify the engine (which I plan to do slightly) there will be merge conflicts in the future.
      This would make upgrades even harder, but I don't see a way around it.
      Apart from maybe proposing the features to be merged into the main Galaxy repo, although they may be too specific.
Thank you,

Regards,


 Vlad Visan Ingénieur de Recherche - OSUG

Bureau 106 OSUG D