OLD-3 APC (unidentified) 192.168.0.240 · dead, no HTTP and no SNMP response on 2026-09-29, replace with SRT-3
PDU rails cross racks: Rack 1 gear sits on rails A and C, Rack 3 gear on C and D. A rack's measured figure below is therefore the output of the UPS standing in that rack, not the draw of the equipment pictured in it. Treat it as UPS load, not rack load.
The three SRT2200 units are racked and on the network but carry no load yet: all three report 0 percent. Rack 3 has no Symmetra and its old 0.240 UPS answers nothing on HTTP or SNMP, so rack 3 is the one rack with no measured protection today. Its estimate of about 2,000W exceeds the 1,980W an SRT2200 can carry, which is why its feed is a decision and not a plan.
Power reroute — the order Andrew works in
Why offload first: SYM-1 fails to bypass on utility loss, which means everything plugged into it goes OFFLINE at that moment. We cannot test or risk that while live gear is on it. So every device must come off SYM-1 and onto the healthy SRT2200s BEFORE anything is tested. Only once SYM-1 is completely empty is it safe to cut its utility input and confirm the fault. This reroute is the prerequisite, not the test.
PRE-STEP for Andrew: confirm which UPS feeds each PDU rail (A/B/C/D). We know the outlets each device uses but not which UPS each rail plugs into. Not a blocker for step 1 below (that moves cords TO the SRTs), but needed before calling any device "redundant".
Move ONE power cord of each of the five cluster nodes into SRT-1: JMFPMX02 (3.232), JMFPMX03 (3.199), JMFPMX04 (0.232), JMFPMX05 (3.190), JMFPMX06 (3.200). Five cords, two outlets to spare. Leave each node's OTHER cord where it is. Measured draw of all five is 922W, so SRT-1 alone carries them at 47 percent. After this, every node has a feed on a healthy UPS regardless of what SYM-1 does.
SRT-2 (192.168.3.251) - cover the single-cord gear
The switches and firewalls are single-PSU, so a Symmetra bypass drops them outright. Put the two core HP switches (0.210, 0.212), both firewalls (0.194, 0.190), the Mini NG switch and one Syn NAS cord onto SRT-2. Move SRT-2 into Rack 1 first, or run cords from Rack 2.
Second cords off SYM-1
Move each node's remaining Symmetra-fed cord so no critical Rack 1 load sits on SYM-1. Confirm with the PRE-STEP rail map which cords those are.
Transfer-test SYM-1 at zero load
With SYM-1 carrying nothing, cut its utility input and watch the transfer. Correct behavior: it switches to battery/inverter and keeps its output live. The fault we are chasing: it drops to bypass, so with no utility the output goes dead. Zero load makes this safe to trigger. If it fails to bypass, the transfer/inverter path or the SYBT2 batteries are the cause and the unit is replaced or rebuilt; nothing was at risk during the test.
Wall mounted
Device
IP
Role
Watts
PDU outlets
Note
Main HP Switch
192.168.0.201
switch
1050
C22
THE BACKBONE SWITCH. Also answers as 192.168.0.1, 192.168.2.1 and 192.168.3.1, one gateway IP per subnet, which is why those addresses looked like separate devices. HPE, ssh and http open, does not answer SNMP with the public community. No credentials on file.
HP Switch
192.168.0.147
switch
630
A23
Dell S4048 (10G)
192.168.0.209
switch
300
10G server aggregation in the East building closet, not the server room. OS9, admin login. Nodes 01-05 have LACP port-channels here; JMFPMX06 is not yet patched in.
HP Switch
192.168.0.208
switch
630
D3
Lobby Ubiquiti
ap
D22
Loft Ubiquiti
ap
A22
Kitchen Ubiquiti
ap
D20
Alarm
other
D3
DVR and monitor
other
D4
PDU rails
Rail
IP
Model
Load
MAC
Location
Note
A
192.168.0.249
AP7932
9.3A
00:c0:b7:cc:ff:66
Rack 3, right
back on the network 2026-09-30 after Andrew reset it; 24 outlets, 1116W at 120V, state normal
B
192.168.0.250
AP7932
3.9A
00:c0:b7:cc:fe:e2
Rack 3, left
back on the network 2026-09-30 after Andrew reset it; 24 outlets, 468W at 120V, state normal. All four rails now measurable.
C
192.168.0.251
AP7932
9.1A
00:c0:b7:cc:fd:88
Rack 2, right
measured 2026-09-30, 1092W at 120V, state normal
D
192.168.0.253
AP7932
8.1A
00:c0:b7:cc:fa:4b
Rack 1, right
measured 2026-09-30, 972W at 120V, state normal
All devices
Rack
U
Device
IP
Role
Nameplate
Measured
PDU outlets
Note
Rack 1
U26
HP Switch
192.168.0.210
switch
630
D11
Rack 1
U25
HP Switch
192.168.0.212
switch
630
D15
Rack 1
U24 to U23
JMFPMX02
192.168.3.232
vm-host
700
149
D24, C21
PowerEdge R730, web
Rack 1
U22
Fiber / Comcast
other
D1, D14
ISP handoff
Rack 1
U21
Firewall
192.168.0.194
firewall
D23
drawing showed C9 too; Andrew reassigned C9 to Fiber 2026-09-30, confirm firewall second feed
Rack 1
U20
Mini NG Switch
switch
D12
Rack 1
U19
F FW
192.168.0.190
firewall
D6
Rack 1
U18 to U17
JMFPMX03
192.168.3.199
vm-host
750
188
D19, C24
PowerEdge R740, hosts JMFDC2 + Exchange
Rack 1
U16 to U15
JMFPMX04
192.168.0.232
vm-host
750
172
A14, C16
PowerEdge R730
Rack 1
U14 to U13
JMFPMX06
192.168.3.200
vm-host
750
148
A19, C2
PowerEdge R740, joined 2026-09-21
Rack 1
U12 to U11
JMFPMX05
192.168.3.190
vm-host
1400
265
C5, A16
PowerEdge R750
Rack 1
U10 to U9
Syn NAS
192.168.0.196
nas
500
C10, D7
Synology 1 of 3, replica NAS, PBS DS001 datastore
Rack 1
U8 to U7
SRT2200
192.168.3.250
ups
SRT-1, no load yet
Rack 1
U6 to U1
Symmetra RM 6000
192.168.0.248
ups
SYM-1, 42 percent, 19 min
Rack 2
U13
HP Switch
192.168.0.211
switch
630
C20
Rack 2
U12 to U11
R520 VM
192.168.3.196
server
750
B7, A24
Rack 2
U10 to U9
R710 VM
192.168.0.227
server
600
A18, C11
Rack 2
U8 to U7
SRT2200
192.168.3.251
ups
SRT-2, no load yet
Rack 2
U6 to U1
Symmetra RM 6000
192.168.0.239
ups
SYM-2, 29 percent, 28 min
Rack 3
U25
Old Sage
other
A13
IDENTITY UNKNOWN. The drawing gave it 192.168.0.1, but Greg confirms 0.1, 2.1 and 3.1 are all gateway IPs on the 192.168.0.201 backbone switch, so that address is not this device. Whatever occupies this U is unidentified.
Rack 3
U24
Backup Main HP Switch
192.168.0.201
switch
460
A12, B24
Rack 3
U23 to U22
R740
192.168.3.241
server
1100
90
A21, B19
Rack 3
U21 to U20
JMFXDATA
192.168.0.197
nas
550
A17, B5
Synology 2 of 3, office NAS, git origin
Rack 3
U19 to U18
R740
192.168.3.226
server
750
115
A1, B9
Rack 3
U17 to U16
HP ProLiant
192.168.0.221
server
400
A11, C7
POWERED OFF and LOCATION UNCONFIRMED: Greg does not believe this is in the rack and is getting details. Kept from the drawing until someone looks. Do not trust the U position or the A11/C7 outlets.
Rack 3
U15 to U14
JMFXDATA2
192.168.0.200
nas
500
A20, C15
Synology 3 of 3
Rack 3
U13 to U12
Proxmox
192.168.0.54
vm-host
750
D13
Rack 3
U11
R610 (Dave)
192.168.0.84
server
700
A7, B1
POWERED OFF, no ping 2026-09-29. Greg is getting details on this one.