I upgraded one of our mock stores and looks like srxpfe is not running do to an issue with space. When looking at vmhost, we see the following, when running df -h:

/dev/sda8 548M 536M 0 100% /var/log <----- file large stopping srxpfe from running since no space

This is an issue on Junos code prior to 22.4, which did not have a log rotation feature for some of the directories on vmhost. One such directory, ovs-vswitchd.log can grow to a huge size, and when this occurs, it can run the partition to 100% capacity. When the /var/log partition gets full, this has an impact on the srxpfe process, which is responsible for all Layer 3 communications on the NFX (including ge-1/0/0, ge-1/0/1, etc ... ) and will affect all srxpfe and flowd traffic.

This issue has resolved in Junos 22.4R2 and higher. This is also documented in https://prsearch.juniper.net/problemreport/PR1721278