You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Bug: RocksDB Memory Consumption Unbounded, Leading to Excessive RAM Usage
Description
The Taraxa node, with built in with RocksDB, does not appear to have memory consumption limits. This results in RocksDB potentially consuming all available system memory, growing to over 70GB in some observed cases. This behavior makes the node unstable and can lead to server crashes. We need to expose RocksDB configuration parameters related to memory management to allow operators to control memory usage.
Steps to Reproduce
Run a mainnet registered Taraxa node (v1.12.7) on a server with 64-128GB of RAM.
Allow the node to run for approximately two days.
Expected Result
The node should consume a reasonable and stable amount of RAM, leaving sufficient memory for other system processes. The memory usage should be configurable and predictable.
Actual Results
The node consumes all available RAM, in some cases exceeding 70GB. This excessive memory usage leads to performance degradation and potential server instability.
Environment
Node version
v1.12.7
Operating System details.
docker container
The text was updated successfully, but these errors were encountered:
Bug Report
Bug: RocksDB Memory Consumption Unbounded, Leading to Excessive RAM Usage
Description
The Taraxa node, with built in with RocksDB, does not appear to have memory consumption limits. This results in RocksDB potentially consuming all available system memory, growing to over 70GB in some observed cases. This behavior makes the node unstable and can lead to server crashes. We need to expose RocksDB configuration parameters related to memory management to allow operators to control memory usage.
Steps to Reproduce
Expected Result
The node should consume a reasonable and stable amount of RAM, leaving sufficient memory for other system processes. The memory usage should be configurable and predictable.
Actual Results
The node consumes all available RAM, in some cases exceeding 70GB. This excessive memory usage leads to performance degradation and potential server instability.
Environment
Node version
v1.12.7
Operating System details.
docker container
The text was updated successfully, but these errors were encountered: