Panic occurs when storage.type filesystem

[k-kinzal]

I got go panic when I set storage.type filesystem on fluent bit v1.4.2 and fluent bit go s3 v0.7.0.

panic: reflect: call of reflect.Value.Index on int64 Value

goroutine 17 [running, locked to thread]:
reflect.Value.Index(0x7f1a281177e0, 0xc0002fb0c0, 0x86, 0x0, 0x0, 0xcd8, 0x1)
	/usr/local/go/src/reflect/value.go:956 +0x1b1
github.com/cosmo0920/fluent-bit-go-s3/vendor/github.com/fluent/fluent-bit-go/output.GetRecord(0xc000302b30, 0x3, 0x3, 0xc0000ccd80, 0xcd8)
	/go/src/github.com/cosmo0920/fluent-bit-go-s3/vendor/github.com/fluent/fluent-bit-go/output/decoder.go:80 +0x106
main.(*fluentPlugin).GetRecord(0x7f1a287fbda0, 0xc000302b30, 0x627, 0xc0005cae00, 0x6b0, 0x7f1a27ea4029)
	/go/src/github.com/cosmo0920/fluent-bit-go-s3/out_s3.go:68 +0x2d
main.FLBPluginFlushCtx(0x0, 0x7f1a26ae10b2, 0x7f1900000f4e, 0x7f1a28d35460, 0x28)
	/go/src/github.com/cosmo0920/fluent-bit-go-s3/out_s3.go:306 +0x109
main._cgoexpwrap_6036375e9d10_FLBPluginFlushCtx(0x0, 0x7f1a26ae10b2, 0xf4e, 0x7f1a28d35460, 0x0)
	_cgo_gotypes.go:88 +0x49

This doesn't always happen and I don't know how to reproduce it. It seems to have a high probability of occurring when the fluent bit is repeatedly restarted.

[k-kinzal]

https://github.com/fluent/fluent-bit-go/blob/master/output/decoder.go#L79-L81

if slice.Kind(m) != reflect.Slice {
	return -1, 0, nil
}
slice := reflect.ValueOf(m)
t := slice.Index(0).Interface()
data := slice.Index(1)

This is a suggestion. It might be better to check that it is slice to be type safe.

[yteraoka]

same here. amazon/aws-for-fluent-bit:2.3.1, FluentBit 1.4.5

panic: reflect: call of reflect.Value.Index on int64 Value

goroutine 17 [running, locked to thread]:
reflect.Value.Index(0x7f500bc36c00, 0x1c000991550, 0x86, 0x0, 0x0, 0x1c0014d8840, 0x194)
	/usr/local/go/src/reflect/value.go:942 +0x1d2
github.com/fluent/fluent-bit-go/output.GetRecord(0x1c000dccc70, 0x1c0000a9d80, 0x80, 0x1c0014d8840, 0xd2f269)
	/go/pkg/mod/github.com/fluent/[email protected]/output/decoder.go:80 +0x10b
main.FLBPluginFlushCtx(0x0, 0x7f5011e7a098, 0x1c000018f68, 0x7f4ff594b920, 0x28)
	/cloudwatch/fluent-bit-cloudwatch.go:134 +0x1c9
main._cgoexpwrap_47a790186e23_FLBPluginFlushCtx(0x0, 0x7f5011e7a098, 0x76732e7400018f68, 0x7f4ff594b920, 0xb96c61636f6c2e72)
	_cgo_gotypes.go:88 +0x49

[SERVICE]
    Parsers_File  /fluent-bit/parsers/parsers.conf
    Parsers_File  /fluent-bit/etc/parsers_file.conf
    HTTP_Server   On
    HTTP_Listen   0.0.0.0
    HTTP_Port     ${FLUENT_HTTP_PORT}
    storage.path  /var/log/fluentbit
    storage.backlog.mem_limit 5M

[INPUT]
    Name              tail
    Tag               kube.*
    Path              /var/log/containers/*.log
    Parser            docker
    DB                /var/log/flb_kube.db
    Mem_Buf_Limit     1MB
    Skip_Long_Lines   On
    Refresh_Interval  10
    Docker_Mode       On
    storage.type      filesystem

[FILTER]
    Name                 kubernetes
    Match                kube.*
    Merge_Log            On
    Merge_Log_Key        data
    K8S-Logging.Parser   On
    K8S-Logging.Exclude  On
    Labels               On

[OUTPUT]
    Name               cloudwatch
    Match              kube.*
    region             ${FLUENT_CLOUDWATCH_REGION}
    log_group_name     /eks/xxx/${FLUENT_CLOUDWATCH_ENV}/k8s
    log_stream_prefix  fluentbit-

[edsiper]

@fluent/contributors

[PettitWesley]

Assigned this to me because I am a member of this repo; there are a few folks on my team who help maintain the AWS go plugins. I'll make sure myself or one of them spends some time troubleshooting this issue this month.

[ghost]

Note identical failure but completely different circumstances. Might shed some light: https://github.com/fluent/fluent-bit-go/issues/34