[Bug]: When running with the TDE, we saw that error states were not getting reported up the chain

Description

When testing k8s with the hugepages issue on np04-srv-031, we saw the following

2026-Feb-25 13:59:22,857 LOG [std::unique_ptr<rte_mempool> dunedaq::dpdklibs::ealutils::get_mempool(const std::string&, int, int, int, int) at /tmp/root/spack-stage/spack-stage-dpdklibs-NFD_DEV_260210_A9-jbbjrl74253o2mx5gd7dpk65oj2nphxg/spack-src/src/EALSetup.cpp:263] get_mempool with: NUM_MBUFS = 16384 | MBUF_CACHE_SIZE = 0 | data_room_size = 16384 | SOCKET_ID = 0
EAL: Error - exiting with code: 1
  Cause: ERROR: Cannot create rte_mempool!

This happened early on and the error was not propagated up the chain

The state was not reported as in error. This is the bug.

Steps to reproduce

Set up a nightly with the TDE readout on np04-srv-031
Change the hugepages config
Run the session and attempt to conf

Version and environment

NFD_DEV_260211_A9

Impact

Small/Isolated

Anything else?

No response

Start at src/EALSetup.cpp:263 and inspect get_mempool, then follow how its failure is handled during the TDE readout session. Reproduce with the hugepages configuration on np04-srv-031 and confirm that the mempool creation error reaches the session state, which should be reported as an error rather than continuing.