This task will setup a new backup system in eqiad, backup1001.

This host will use a spare server wmf4750, and also use parts from helium.

Helium currently has an H800 external raid array controller, plus a MD1200 disk shelf (helium-array). We'll need the H800 removed from helium, and installed into wmf4750.

Downtime/decom of helium will need to be scheduled with @akosiaris, as soon as we yank out the H800 the old system is useless.

• - apply hostname label to wmf4750 as backup1001.
• - unrack backup1001 (d3-eqiad), prepare to move it into the space currently occupied by helium (a8-eqiad).
• - @Cmjohnson to coordinate with @akosiaris on when to decom/yank parts from helium.
• - migrate helium out of its current rackspace, move where there is room to wipe disks (doesn't matter where) and place backup1001 in helium's old spot (easier to move the 1U server than move the MD1200 2U disk shelf and find space for it in the same rack where backup1001 currently resides.)
• - @Cmjohnson pulls H800 controller out of helium, connects disk shelves, ensures disks are detected and no data loss happens (it should allow the config, stored on the H800, to simply build and use the existing raid array in its shelf.)
• - OS installation of backup1001, on its internal disks, NOT the md1200 disks.
• - handoff to @akosiaris for service deployment of backup service