Description of problem:

Single log line produced by the container as:

2024-08-28T12:14:11.381+0200: [GC (Allocation Failure) 2024-08-28T12:14:11.381+0200: [ParNew: 22229K->401K(27328K), 0.0042127 secs] 132842K->111021K(185064K), 0.0042535 secs] [Times: user=0.00 sys=0.00, real=0.01 secs]  

Split in several by CRIO (it's also printed the end of line for better observability for troubleshooting)

2024-08-29T09:42:35.446720302+00:00 stdout P 2024-08-28T12:14:11.381+0200: [GC (Allocation Failure) 2024-08-28T12:14:11.381+0200: [ParNew: 22229K->401K(27328K), 0.0042127 secs] 132$
2024-08-29T09:42:35.446756250+00:00 stdout F 842K->111021K(185064K), 0.0042535 secs] [Times: user=0.00 sys=0.00, real=0.01 secs] $

And reassembled vy the collector the original log line, it's marked ar "_partial: true". This is how it looks like in Loki side:

_partial: true
...
message: 2024-08-28T12:14:11.381+0200: [GC (Allocation Failure) 2024-08-28T12:14:11.381+0200: [ParNew: 22229K->401K(27328K), 0.0042127 secs] 132842K->111021K(185064K), 0.0042535 secs] [Times: user=0.00 sys=0.00, real=0.01 secs] 

Version-Release number of selected component (if applicable):

$ oc get csv -n openshift-logging|grep logging
cluster-logging.v5.9.5                  Red Hat OpenShift Logging   5.9.5                  cluster-logging.v5.9.4                         Succeeded

How reproducible:

Always

Steps to Reproduce:

1. Deploy CLO and not needed to create the CLF
2.  Create a pod as hello-node: `kubectl create deployment hello-node --image=registry.k8s.io/e2e-test-images/agnhost:2.43 – /agnhost serve-hostname`
3.  Upload to this "hello-node" pod the file attached using `oc cp <log file> > <hello-node pod>:/tmp`
4. `oc rsh <hello-node pod>` and generate the logs `cat <logfile > /proc/1/fd/1`
5. Running `vector tap --inputs-of output_default_loki_apps` in the collector pod running in the same node that the `hello-node` pod

NOTE

Actual results:

Single log line produced by the container split by CRIO in several is marked by Vector   as  `_partial: true` even when the original log line was reassembled entirely by the collector:

_partial: true ... message: 2024-08-28T12:14:11.381+0200: [GC (Allocation Failure) 2024-08-28T12:14:11.381+0200: [ParNew: 22229K->401K(27328K), 0.0042127 secs] 132842K->111021K(185064K), 0.0042535 secs] [Times: user=0.00 sys=0.00, real=0.01 secs]  

Expected results:

It shouldn't be labelled as "_partial: true" as the entire log was reassembled by Vector and the output event contains all the original data.

Additional note:

Screenshot about how it looks like in the Loki side is attached