Docker deployment: MEMORY_SIZE settings in confignode-env.sh and datanode-env.sh not effective, leading to continuous memory increase during compaction recovery (versions 1.3.2 and 2.0.5)

[MM-X-C]

In a Docker-based standalone deployment of IoTDB, setting MEMORY_SIZE (e.g., MEMORY_SIZE=4G) in iotdb-env.sh does not take effect. During startup, the compaction recovery process (specifically "recovering sequence inner compaction") causes the container memory usage to continuously increase without bound, eventually leading to OOM or high resource exhaustion. This prevents the RPC service from becoming available (connection refused on port 6667). The issue persists across versions 1.3.2 and 2.0.5, with no visible impact from the JVM heap configuration.
This occurs in standalone mode. No errors are logged beyond the repeated recovery messages, and basic resource limits in docker-compose.yml (e.g., limits: memory: 6G) do not mitigate the growth.

[github-actions]

Hi, this is your first issue in IoTDB project. Thanks for your report. Welcome to join the community!

[CritasWang]

Docker runtime needs to be specified in this way

environment
- IOTDB_JMX_OPTS=-Xms4G -Xmx4G -XX:MaxDirectMemorySize=1G
- CONFIGNODE_JMX_OPTS=-Xms1G -Xmx1G -XX:MaxDirectMemorySize=256M