Milvus system logs in kubenetes environment

[nileshsironja]

I'm looking to get the milvus system logs in K8s environment as shown below.

I've set the path in milvus.yaml but still I don't see any system level logs. I also tried to get the logs in local path, that too not generating any system logs. Logs generated through export-milvus-log.sh are for pods control path level.

log:
  file:
    maxAge: 10
    maxBackups: 20
    maxSize: 300
    rootPath: /milvus/
  format: text
  level: debug
  stdout: true

Here is logs which I get in standalone milvus setup:

I20250305 18:04:41.091521 357627 index_factory.cc:55] [KNOWHERE][Create][milvus] use key STARLING_fp32 to create knowhere index STARLING with version 6
Opening bin file slice_prefix_ids.bin...
I20250305 18:04:41.292883 357627 index.cpp:1579] [KNOWHERE][link][milvus] Completed Pass 1 of data using L=100 and alpha=1.2. Stats: search+prune_time=0.0461611s, inter_time=0.00242495s, inter_count=1
I20250305 18:04:41.292940 357627 index.cpp:1775] [KNOWHERE][build][milvus] Index built with degree: max:34  avg:12.8316  min:3  count(deg<2):0
Indexing time: 0.111176
I20250305 18:04:41.294811 357627 aux_utils.cpp:1235] [KNOWHERE][build_disk_index][milvus] Starting index build: R=56 L=100 Query RAM budget: 0.00232825(GiB) Indexing ram budget: 59.8729(GiB)

[yhmo]

milvus standalone in k8s? You can use kebuctl exec to enter the container, there should be a file "standalone.log" under the path /milvus

[nileshsironja]

I'm running milvus clustered one.
And I found that I need to enable the flag for disk index in queryNode.

set queryNode.enableDisk to true.

[yhmo]

queryNode.enableDisk is for DISKANN index, not for log.
https://milvus.io/docs/disk_index.md#Prerequisites