Laravel Horizon not trimming jobs?

[mravdel]

I use Laravel Forge with AWS (EC2, Ubuntu) and Laravel Horizon to run jobs.

My Redis memory consumption is only growing in size, it is already almost 7GB, which is insane for the amount of jobs I store
Queue size is about 9000 jobs, it was way more a month ago, but it did take a lot less space. Every job is very small in size and highly optimised.

Seems like Redis database only growing and never delete old keys. is is Horizon problem?
My trim settings from horizon.php:

'trim' => [
        'recent' => 60,
        'pending' => 43200,
        'completed' => 60,
        'recent_failed' => 10080,
        'failed' => 10080,
        'monitored' => 10080,
    ],

INFO MEMORY From Redis:

# Memory
used_memory:7215103344
used_memory_human:6.72G
used_memory_rss:7296204800
used_memory_rss_human:6.80G
used_memory_peak:7289004080
used_memory_peak_human:6.79G
used_memory_peak_perc:98.99%
used_memory_overhead:86652434
used_memory_startup:796224
used_memory_dataset:7128450910
used_memory_dataset_perc:98.81%
allocator_allocated:7217543192
allocator_active:7347945472
allocator_resident:7378243584
total_system_memory:16596881408
total_system_memory_human:15.46G
used_memory_lua:51200
used_memory_lua_human:50.00K
used_memory_scripts:2832
used_memory_scripts_human:2.77K
number_of_cached_scripts:6
maxmemory:0
maxmemory_human:0B
maxmemory_policy:noeviction
allocator_frag_ratio:1.02
allocator_frag_bytes:130402280
allocator_rss_ratio:1.00
allocator_rss_bytes:30298112
rss_overhead_ratio:0.99
rss_overhead_bytes:-82038784
mem_fragmentation_ratio:1.01
mem_fragmentation_bytes:81060576
mem_not_counted_for_evict:0
mem_replication_backlog:0
mem_clients_slaves:0
mem_clients_normal:487522
mem_aof_buffer:0
mem_allocator:jemalloc-5.2.1
active_defrag_running:0
lazyfree_pending_objects:0