I believe that oci-flexvolume-driver outputs a very high amount of logs, volume of which I would consider highly abnormal.
I think it's on average around 800-1500 log lines per minute per container, which is a lot.
Actually, oci-flexvolume-driver is by far the most "verbose" pod in my clusters, even on the default INFO level logging.
The actual contents of the logs are quite useless too (at a glance) - basically stuff like this many times per second
2020-12-31T16:18:58.758Z INFO client/utils.go:94 OCI using read rate limit configuration: QPS=20, bucket=5 {"pid": 26571, "version": "", "build": "", "command": "init"}
2020-12-31T16:18:58.758Z INFO client/utils.go:98 OCI using write rate limit configuration: QPS=20, bucket=5 {"pid": 26571, "version": "", "build": "", "command": "init"}
2020-12-31T16:18:58.775Z INFO metadata/instance_metadata.go:87 Metadata endpoint /opc/v2/instance/ returned response successfully
2020-12-31T16:18:58.776Z INFO metadata/instance_metadata.go:87 Metadata endpoint /opc/v2/instance/ returned response successfully
What you expected to happen?
I would be happier if oci-flexvolume-driver didn't output that many logs, or at least allowed me to reduce this verbosity.
How to reproduce it (as minimally and precisely as possible)?
I think this is the standard behavior of oci-flexvolume-driver on any environment, so it should be easy to reproduce
Anything else we need to know?
I think it's also worth investigating why the "init" operation happens so frequently - I don't have much knowledge into this stuff, but it seems quite unnatural.
mrunalpagnis
commented
3 years ago
Is this a BUG REPORT or FEATURE REQUEST?
BUG REPORT
Versions
CCM Version:
0.10.0
Environment:
kubectl version): 1.15.6uname -a): 4.14.35-1844.1.3.el7uek.x86_64What happened?
I believe that oci-flexvolume-driver outputs a very high amount of logs, volume of which I would consider highly abnormal. I think it's on average around 800-1500 log lines per minute per container, which is a lot. Actually, oci-flexvolume-driver is by far the most "verbose" pod in my clusters, even on the default INFO level logging.
The actual contents of the logs are quite useless too (at a glance) - basically stuff like this many times per second
What you expected to happen?
I would be happier if oci-flexvolume-driver didn't output that many logs, or at least allowed me to reduce this verbosity.
How to reproduce it (as minimally and precisely as possible)?
I think this is the standard behavior of oci-flexvolume-driver on any environment, so it should be easy to reproduce
Anything else we need to know?
I think it's also worth investigating why the "init" operation happens so frequently - I don't have much knowledge into this stuff, but it seems quite unnatural.