I've a shitload of UPSs -- most in the office.

Power *tends* to be pretty reliable, here, so they aren't
used for their "backup" capability. However, we do tend to
see "switching transients", predictably, at certain times
of the day. So, having a "second" of carryover is often
helpful (on the smaller machines; the larger ones have big
enough power supplies that they can bridge these momentary
outages)

Instead, they tend to be "outlet multipliers" in much the
same way as outlet strips might be (though a better form
factor, given the layout of my kit). They let me connect
all of the devices for a host to the UPS associated with that
host. Then, power them all on/off at the same time.

But, most importantly, it lets me get a feel for how much
of a load I have on the mains -- so I don't carelessly
add some new load that blows a breaker!

Replacing batteries is a costly -- and time consuming -- exercise.
A single "shared" battery would be the preferred solution. But,
not possible with COTS UPSs (nor the associated IR drops).

OTOH, I *could* plug more than one UPS into a single (or few)
"master" UPS that effectively takes on the role of the mains
connection. Then, I just have to keep that/those UPS running
("properly") to gain the benefits of bridging those brief glitches.

Any downsides to this? If I remove the batteries from the
"slave" UPSs (i.e., don't replace them when they fail), then
I can avoid that maintenance issue but still retain the other
advantages of the UPS-per-workstation.

Or, maybe put some trivial "power source" in their place that
is just enough to satisfy the UPSs' expectations of a "battery"?

[I'd still have to put up with the "connect battery" messages
but maybe I'll hack the firmware to "fix" that issue -- and,
stop the silly scrolling of "screens" so I can just see the
data of interest, constantly.]

[[UPS manufacturers seem to have shitty/buggy/unfriendly software]]

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)