@ayounsi i think this is the same as what to had on https://phabricator.wikimedia.org/T294009

I create ticket # 1-235341265861 requesting Equinix to check the breaker on the feed where PEM0 is connected.

Technician Note
Equinix Support , Jun/12/2024 22:28
The site has investigated customer equipment 2016250 Juniper in cabinet 504. All power indicators are green. The only alarm in red was located in the front of the device. Please see the attached images.

Reopen Note
Papaul Tshibamba , Jun/12/2024 23:19
Thank you for checking this yes indeed all the power indicators are green but we are not getting enough power on PEM 0 that explain the red alarm located in front of the device and that is what we are getting also on our end in our monitoring system. like i mentioned we had this issue before and the fix was somewhere at the breaker side on the feed coming to our cage. Reference case number 1-214270167279. Thank you
~~~
cr2-eqdfw> show chassis alarms
1 alarms currently active
Alarm time Class Description
2024-06-07 01:48:07 UTC Major PEM 0 Input Voltage Out Of Range

We Will be going on site this Monday, June 17th at 11am to work with Equinix team on fixing this issue. @cmooney will be depooling the site.

Mentioned in SAL (#wikimedia-operations) [2024-06-17T15:20:40Z] <topranks> draining transport circuits in/out of eqdfw in advance of router power-supply work/upgrade T366864

Icinga downtime and Alertmanager silence (ID=8c9e583f-3e7e-41d4-ab2f-0f862f085c35) set by cmooney@cumin1002 for 1:30:00 on 2 host(s) and their services with reason: JunOS upgrade and PSU swap on cr2-eqdfw

cr2-eqdfw,cr2-eqdfw IPv6

Mentioned in SAL (#wikimedia-operations) [2024-06-17T15:39:31Z] <topranks> deactivate Tranist and peering sessions on cr2-eqdfw in advance of power-supply change T366864

Icinga downtime and Alertmanager silence (ID=a784f421-3f86-49b7-a9f6-15e53a3bc540) set by cmooney@cumin1002 for 1:00:00 on 5 host(s) and their services with reason: JunOS upgrade and PSU swap on cr2-eqdfw

cr[1-2]-codfw,cr2-drmrs,cr2-esams,cr2-magru

Icinga downtime and Alertmanager silence (ID=2d8f617e-3efd-4a08-88cb-75ab38c0cc68) set by cmooney@cumin1002 for 0:40:00 on 5 host(s) and their services with reason: JunOS upgrade and PSU swap on cr2-eqdfw

cr[1-2]-codfw,cr2-drmrs,cr2-esams,cr2-magru

Icinga downtime and Alertmanager silence (ID=f6da6e6c-9be7-4e8c-9beb-a7f1139359f6) set by cmooney@cumin1002 for 0:40:00 on 2 host(s) and their services with reason: JunOS upgrade and PSU swap on cr2-eqdfw

cr2-eqdfw,cr2-eqdfw IPv6

PEM0 was connected to ps1 with blue tag and PEM1 was connected to PS2 with red tag.
moved PEM0 to ps2 error clear
moved PEM1 to PS1 error moved from PEM0 to PEM1
power down the router to swap out ps1 and ps2 from the feed, PEM0 didn't come back up.
Re-seated PEM0 same issue

Dear PAPAUL TSHIBAMBA,

Thank you for contacting Juniper Networks Global Support.

Case 2024-0617-183459 with Priority of P2 - High has been CREATED by you or a Juniper Agent to track issue as described below.

To update this Case click here to go to Juniper Support Portal Only registered users have access. Click here to register.

• Case Details *

Case Number: 2024-0617-183459
Customer Tracking Number:
Create Date/Time in UTC: JUN 17 2024 20:47:16

Synopsis: RMA request for the Power Supply

Description:

RMA request for the Power Supply

Power supply will be shipped out today

Hello Papaul,

Thank you for the information provided.

This is the information for the replacement of the faulty PEM with serial number 1F188120554. The RMA ID is # R200519348.

F

Your replacement part associated with RMA R200519348 Item # 100 has been successfully shipped. Details of which are provided below.
Tracking URL: https://wwwapps.ups.com/WebTracking/track?track=yes&trackNums=1Z7AF3880172299886

We replace PEM0 today with the new one that was send to us from Juniper same error. When we move PEM0 to the PDU where PEM1 is plugged in the error clear. So I open ticker # 1-235774184519 with Equinix requesting to check.Equinix repaced the Breaker same problem, PDU and receptacle same problem. So the engineer told us that they will move us on another socket to give us about 1 day for that.