Skip to content

[BUG] used_memory is much larger than used_memory_rss #6302

Open
@become-nice

Description

@become-nice

Describe the bug
I now have a Redis instance loaded with RediSearch. The used_memory usage is ten times that of used_memory_rss, but the amount of data I have is very small, even less than used_memory_rss.

127.0.0.1:6379> INFO memory
# Memory
used_memory:5854204632
used_memory_human:5.45G
used_memory_rss:560148480
used_memory_rss_human:534.20M
used_memory_peak:5871460488
used_memory_peak_human:5.47G
used_memory_peak_perc:99.71%
used_memory_overhead:142966344
used_memory_startup:4003088
used_memory_dataset:5711238288
used_memory_dataset_perc:97.62%
allocator_allocated:5854546192
allocator_active:5902098432
allocator_resident:5922619392
total_system_memory:8326955008
total_system_memory_human:7.76G
used_memory_lua:63488
used_memory_lua_human:62.00K
used_memory_scripts:3576
used_memory_scripts_human:3.49K
number_of_cached_scripts:7
maxmemory:6442450944
maxmemory_human:6.00G
maxmemory_policy:noeviction
allocator_frag_ratio:1.01
allocator_frag_bytes:47552240
allocator_rss_ratio:1.00
allocator_rss_bytes:20520960
rss_overhead_ratio:0.09
rss_overhead_bytes:-5362470912
mem_fragmentation_ratio:0.10
mem_fragmentation_bytes:-5294015128
mem_not_counted_for_evict:0
mem_replication_backlog:134217728
mem_clients_slaves:20512
mem_clients_normal:2748552
mem_aof_buffer:0
mem_allocator:jemalloc-5.1.0
active_defrag_running:0
lazyfree_pending_objects:0
lazyfreed_objects:0

To Reproduce
Not sure unfortunately, seems to happen randomly.

Expected behavior
Value to be accurate

Environment (please complete the following information):

  • OS: debian 10
  • CPU model
  • Version/branch [2.6.13]

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions