On all MX Series and PTX Series routers running Junos OS, when the periodic packet manager (ppman) process runs short of resource, LACP protocol data units (PDUs) would fail to send.
This causes LACPD_TIMEOUT in the peer, which consequently causes LACP flap in member links and the BFD session on the aggregated Ethernet interface will go down.
The peer can see the LACP timeout by using the command 
user@device>show lacp timeouts
Aggregated interface: ae57
    LACP Interfaces  Current state           Last timeout
      et-2/0/8        Collecting distributing  xxxx-xx-xx 21:07:14 PST (06:06:11 ago). <=======
      et-7/0/6        Collecting distributing  xx-xx-xx 19:37:57 PST (07:35:28 ago)

During this issue, you can see the PPM resource starvation with these commands in the Packet Forwarding Engine shell:  "clear ppm statistics" and "show ppm statistics detail".

This issue might be seen if the following conditions are met:
* On all Junos MX/PTX platforms
* LACP enabled
* BFD on aggregated Ethernet interface
* ppman runs short of resources

Aggregated interfaces would flap causing protocols to flap, child interfaces of the AE bundle will not flap.

> request pfe execute command "show ppm statistics detail" target fpc<> ------> target fpc can be any FPC which has child interfaces of the AE.

Data Scheduler stats

Max time between PPMAN schedules : 715068870 >>>>> This will be more than 3000 seconds [3 PDU cycles of LACP]

Max run time           : 0

Scheduler threshold limit     : 75

Run time threshold limit     : 75

Scheduler threshold limit exceptions: 18

Run time threshold limit exceptions : 0

Enable the option "set protocols lacp ppm quick-processed".

Resolved-In

evo:22.4R3-EVO evo:23.1R2-EVO evo:23.2R1-EVO junos:20.3X75-D36 junos:20.3X75-D440 junos:20.3X75-D46 junos:20.4R3-S10 junos:22.3R2-S2 junos:22.3R3 junos:22.4R2 junos:22.4R3 junos:23.1R2 junos:23.2R1 junos:23.4R2 junos:24.1R2 junos:24.2R1

https://prsearch.juniper.net/PR1706224