With LTS 19 we see during "reboot" of CN96xx that Hardware RNG Entropy Gatherer Daemon usually gets stuck while stopping. This may of may not work, which may be caused the daemon waiting for something before stopping. In any case it stalls the reboot while the issue occurs and results in extremely long reboot cycles. [ **] A stop job is running for Hardware RNG Entropy Gatherer Daemon (1min 12s / 1min 31s) A quick search shows this is a knows issue: [https://bugzilla.redhat.com/show_bug.cgi?id=1690364] Our LTS19 rng-tools 6.7 already includes part of the fix from the initial discussion: [https://github.com/nhorman/rng-tools/commit/62fbff0ac807e07f4f21f58c0f15bfe5e9c8b55e|https://github.com/nhorman/rng-tools/commit/62fbff0ac807e07f4f21f58c0f15bfe5e9c8b55e#diff-125de59086175be28f557034dddaec06] But it appears this is not complete fix and some follow-ups exist: [https://github.com/nhorman/rng-tools/commit/d719ae84d09af63e511dbea5228e34ed837cbfa6] The merge above includes: Allow jitter threads to exit faster [https://github.com/nhorman/rng-tools/commit/3e47faae108df4824531bf9c003cc1c65b7b2842] Daemonize earlier in init [https://github.com/nhorman/rng-tools/commit/cf1475fbdf33d5c3a62099b6b20f2b2017d6de33] Can these fixes from rng-tools 6.8 and later be investigated and backported into our product?