The affected interface seems [0] to be always HundredGigabitEthernetab/0/0.
Not sure if this is an infra issue or if the hoststack tests are not re-initializing the interface properly, but the same tests on 3n-icx are passing, as do the normal trex+vpp tests on 3nb-spr.
In trending, sometimes also the first test fails [1], maybe it depends on which job was running on the testbed before.

[0] https://s3-logs.fd.io/vex-yul-rot-jenkins-1/csit-vpp-perf-report-iterative-2310-3nb-spr/38/log.html.gz#s1-s1-s1-s1-s2-t1-k2-k4-k7-k7
[1] https://s3-logs.fd.io/vex-yul-rot-jenkins-1/csit-vpp-perf-hoststack-daily-master-3nb-spr/87/log.html.gz#s1-s1-s1-s1-s1-t1-k2-k3-k7-k7