Concordium / Concordium/concordium-node
Investigate logging caused by db recovery test script
- Dominant language
- Haskell
- Stars
- 108
- Forks
- 24
- Avg merge
- 2d 4h
- Merged PRs (30d)
- 9
Description
**Bug Description**
Testing db-recovery on Testnet with P6 active revealed unusual logging. While running a script that truncates the blockstate file before restarting the node, the following log appears occasionally:
```
concordium-testnet-node-6.0.4: mmap 4096 bytes at (nil): Cannot allocate memory
concordium-testnet-node-6.0.4: Try specifying an address with +RTS -xm -RTS
timeout: the monitored command dumped core
```
**Steps to Reproduce**
1. Catch up with Testnet node to the final height.
2. Move/copy the db files to the path that doesn't require elevated permissions
3. Change ownership of db files to be your username
4. Run the script that can be found [here](https://github.com/Concordium/concordium-test-data/blob/main/test-scripts/truncate-and-recover.sh)
**Expected Result**
Mentioned logs are not expected to appear. We should try to investigate and understand why do they appear.
**Actual Result**
```
2023-08-23T11:20:14.943060987Z: INFO: Starting up concordium_node version 6.0.4!
2023-08-23T11:20:14.943071552Z: INFO: Application data directory: /home/mh/scripts/db-recovery/concordium-4221332d34e1694168c2a0c0b3fd0f273809612cb13d000d5c2e00e85f50f796/data
2023-08-23T11:20:14.943072760Z: INFO: Application config directory: /home/mh/scripts/db-recovery/concordium-4221332d34e1694168c2a0c0b3fd0f273809612cb13d000d5c2e00e85f50f796/config
2023-08-23T11:20:14.943073303Z: INFO: Network: enabled
2023-08-23T11:20:14.943073786Z: INFO: Log level: info
2023-08-23T11:20:14.994655777Z: INFO: My Node ID is 0000000000000001
2023-08-23T11:20:14.994663405Z: INFO: Listening on 172.17.0.1:8888
2023-08-23T11:20:14.996487665Z: INFO: Starting consensus layer
2023-08-23T11:20:14.996491233Z: INFO: Starting up the consensus thread
2023-08-23T11:20:14.997117648Z: INFO: Starting up the consensus layer
concordium-testnet-node-6.0.4: mmap 4096 bytes at (nil): Cannot allocate memory
concordium-testnet-node-6.0.4: Try specifying an address with +RTS -xm -RTS
timeout: the monitored command dumped core
```
Also worth to mention, this logs appearance is not consistent, sometimes they do appear at the restart but mostly they don't.
It seems that they are more often to show when starting the test (when the script iterates through the loop for the first and second time)
**Versions**
- Software Version: 6.0.4 Testnet
- OS: Ubuntu 22.04.3 LTS
Log attached
[stderr-1.log](https://github.com/Concordium/concordium-node/files/12422028/stderr-1.log)
Contributor guide
Assessment
This issue has not been assessed yet.