[BUG] SYS-PROBE | ERROR | (pkg/security/resolvers/process/resolver_ebpf.go:1335 in newEntryFromProcfsAndSyncKernelMaps) | couldn't marshal proc_cache entry: invalid inode/mountID couple: 0/0

[mikkoc]

Agent Environment

$ kubectl exec -it datadog-agent-l82lv -- agent version
Defaulted container "agent" out of: agent, trace-agent, process-agent, system-probe, init-volume (init), init-config (init), seccomp-setup (init)
Agent 7.63.2 - Commit: 51c6eb811f - Serialization version: v5.0.141 - Go version: go1.23.5

Describe what happened:
We upgraded the Agent from 7.58.2 to 7.63.2.
These kind of errors started appearing in our agent logs:

2025-03-03 15:35:52 UTC | SYS-PROBE | ERROR | (pkg/security/resolvers/process/resolver_ebpf.go:1335 in newEntryFromProcfsAndSyncKernelMaps) | couldn't marshal proc_cache entry: invalid inode/mountID couple: 0/0

Describe what you expected:
No new errors after upgrading.

Steps to reproduce the issue:
Upgrade the DD Agent from 7.58.2 to 7.63.2.

Additional environment details (Operating System, Cloud provider, etc):

• Amazon Linux 2023 - EKS optimised.
• EKS 1.30
• Datadog Helm Chart 3.100.0

[paulcacheux]

Hello 👋 thanks for reporting this. This error message is related to a race that happens when we start the process resolution part of system-probe and is caused by a small race during initialization. Please feel free to ignore this error for now, a fix for it that will be included in agent version 7.65.0.