PMM 2.44.0 reporting doubled swap metrics on restart

[GenLN]

Description

PMM agent on restart for some period around 5m reports false swap usage, swap size, it looks like it gets double value.

Expected Results

correct values for swap usage/size metrics

Actual Results

doubled values for few minutes

Version

pmm server 2.44.0
pmm client 2.44.0

Steps to reproduce

1. run for some time mysql node and pmm agent (docker env), getting correct metrics for swap...
2. down/up containers
3. check metrics again

we used this query to get usage percentage by each node
round(sum by(node_name)(((node_memory_SwapTotal_bytes{} - node_memory_SwapFree_bytes{}) /node_memory_SwapTotal_bytes{}) * 100),0.01)

maybe aggregation here is causing problem with this sum by(), but same query was working very well before...
like after down/up action swap would be free from usage, but we get like usage report over 100% for few minutes...

Relevant logs

Code of Conduct

• I agree to follow Percona Community Code of Conduct

[BupycHuk]

Hi, can you provide more details about your environment?