Errors when profiling application with large debuginfo

[Ilmarii]

Hi,
When trying to profile our Vector deployments, I found that although the samples are displayed normally, there are no function names.
And i found this error in logs:

level=warn name=parca ts=2022-12-01T18:49:13.161878448Z caller=store.go:423 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to validate debug information" err=EOF
level=warn name=parca ts=2022-12-01T18:49:13.162012268Z caller=store.go:435 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to fetch debuginfod file" err="failed to fetch from debuginfod: debug info not found"
level=error name=parca ts=2022-12-01T18:49:13.162494329Z caller=store.go:232 component=debuginfo msg="failed to validate object file" buildid=971415fa19d528ee99a74555d4ce0646039eee59
level=error name=parca ts=2022-12-01T18:49:13.162527249Z caller=logger.go:38 protocol=grpc grpc.component=server grpc.service=parca.debuginfo.v1alpha1.DebugInfoService grpc.method=Upload grpc.method_type=client_stream peer.address=100.96.37.20:60480 grpc.start_time=2022-12-01T18:49:13Z grpc.request.deadline=2022-12-01T18:50:54Z grpc.code=Internal grpc.error="rpc error: code = Internal desc = EOF" grpc.time_ms=0.826 finishedcall=(MISSING)
level=warn name=parca ts=2022-12-01T18:49:13.464684326Z caller=store.go:423 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to validate debug information" err=EOF
level=warn name=parca ts=2022-12-01T18:49:13.464816583Z caller=store.go:435 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to fetch debuginfod file" err="failed to fetch from debuginfod: debug info not found"
level=error name=parca ts=2022-12-01T18:49:13.465469569Z caller=store.go:232 component=debuginfo msg="failed to validate object file" buildid=971415fa19d528ee99a74555d4ce0646039eee59
level=error name=parca ts=2022-12-01T18:49:13.465508493Z caller=logger.go:38 protocol=grpc grpc.component=server grpc.service=parca.debuginfo.v1alpha1.DebugInfoService grpc.method=Upload grpc.method_type=client_stream peer.address=100.96.24.75:37130 grpc.start_time=2022-12-01T18:49:13Z grpc.request.deadline=2022-12-01T18:51:04Z grpc.code=Internal grpc.error="rpc error: code = Internal desc = EOF" grpc.time_ms=0.989 finishedcall=(MISSING)
level=warn name=parca ts=2022-12-01T18:49:13.481096585Z caller=store.go:423 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to validate debug information" err=EOF
level=warn name=parca ts=2022-12-01T18:49:13.481212601Z caller=store.go:435 component=debuginfo buildid=971415fa19d528ee99a74555d4ce0646039eee59 msg="failed to fetch debuginfod file" err="failed to fetch from debuginfod: debug info not found"
level=error name=parca ts=2022-12-01T18:49:13.481889726Z caller=store.go:232 component=debuginfo msg="failed to validate object file" buildid=971415fa19d528ee99a74555d4ce0646039eee59
level=error name=parca ts=2022-12-01T18:49:13.481924931Z caller=logger.go:38 protocol=grpc grpc.component=server grpc.service=parca.debuginfo.v1alpha1.DebugInfoService grpc.method=Upload grpc.method_type=client_stream peer.address=100.96.36.219:60862 grpc.start_time=2022-12-01T18:49:13Z grpc.request.deadline=2022-12-01T18:51:04Z grpc.code=Internal grpc.error="rpc error: code = Internal desc = EOF" grpc.time_ms=0.981 finishedcall=(MISSING)

When I tried to load debuginfo with cli, an error also occurred:

parca-debuginfo upload --store-address=localhost:7070 --insecure /tmp/vector-x86_64-unknown-linux-gnu/bin/vector 
level=error err="1 error occurred:\n\t* failed to upload debug information: send chunk: rpc error: code = Internal desc = EOF\n\n"

Founded workaround is to stop all parca agents, then clear the object store on the parca server in /tmp, and then load the debuginfo with parca-debuginfo.
After these steps, symbolization proceeds normally.
The size of Vector debuginfo is about 2.1 gigabytes. (Just build release with debug symbols enabled.)

Parca version: 0.14
Parca-agent version: 0.10
Parca and Vector deployed in k8s.
Vector: https://github.com/vectordotdev/vector