0

I setup a Redis (version 4.0.6) sentinel cluster in two centos 6 VMs. Both master and slave Redis server has maxmemory set to 10GB and maxmemory_policy as volatile-lru.

The problem is, both servers are taking alot of memory.

Master:

# Memory
used_memory:9939524616
used_memory_human:9.26G
used_memory_rss:11554213888
used_memory_rss_human:10.76G
used_memory_peak:10001440960
used_memory_peak_human:9.31G
used_memory_peak_perc:99.38%
used_memory_overhead:1469809500
used_memory_startup:761776
used_memory_dataset:8469715116
used_memory_dataset_perc:85.22%
total_system_memory:20957556736
total_system_memory_human:19.52G
used_memory_lua:37888
used_memory_lua_human:37.00K
maxmemory:10000000000
maxmemory_human:9.31G
maxmemory_policy:volatile-lru
mem_fragmentation_ratio:1.16
mem_allocator:jemalloc-3.6.0
active_defrag_running:0
lazyfree_pending_objects:0

Slave:

# Memory
used_memory:9930822872
used_memory_human:9.25G
used_memory_rss:12246667264
used_memory_rss_human:11.41G
used_memory_peak:9999999848
used_memory_peak_human:9.31G
used_memory_peak_perc:99.31%
used_memory_overhead:1468959861
used_memory_startup:761792
used_memory_dataset:8461863011
used_memory_dataset_perc:85.21%
total_system_memory:20957556736
total_system_memory_human:19.52G
used_memory_lua:37888
used_memory_lua_human:37.00K
maxmemory:10000000000
maxmemory_human:9.31G
maxmemory_policy:volatile-lru
mem_fragmentation_ratio:1.23
mem_allocator:jemalloc-3.6.0
active_defrag_running:0
lazyfree_pending_objects:0

I did run MEMORY PURGE which clears some of the rss memory but it gets re populated within a few minutes and keeps growing.

Most of my data stored in Redis has expiry time although some keys have no expiry set.

Any suggestion on how I can control the memory consumption or a permanent solution for this issue. Should I just increase the RAM on the VMs? I do run other applications on both of the VMs but they take hardly 100 MB of memory.

1
  • You set maxmemory to 10G so that's what gets used. This is normal expected behavior. Dec 28, 2018 at 15:04

1 Answer 1

0

The memory stats indicate that most of the memory is used by the data so there's nothing really wrong with that.

If fragmentation is an issue, you can try enabling the active defragmentation. Lastly, I'd look into updating the Redis server to the latest v4.x, if not 5.x.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .