panic when using snmp in v1.21.3

[ptxmac]

Relevant telegraf.conf

[[inputs.snmp]]
  agents = [ "tcp://<hidden>"]

  timeout = "30s"

  version = 2

  community = "public"

  [[inputs.snmp.field]]
    oid = "RFC1213-MIB::sysUpTime.0"
    name = "uptime"

  [[inputs.snmp.field]]
    oid = "RFC1213-MIB::sysName.0"
    name = "sysName"
    is_tag = true

  [[inputs.snmp.table]]
    oid = "UCD-SNMP-MIB::laTable"

  [[inputs.snmp.table.field]]
    oid = "UCD-SNMP-MIB::laNames"
    is_tag = true

Logs from Telegraf

2022-02-06T11:29:50Z D! [agent] Initializing plugins
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x18 pc=0x2c77238]

goroutine 1 [running]:
github.com/influxdata/telegraf/internal/snmp.LoadMibsFromPath.func2({0xc000cdf7c0, 0x42}, {0x0, 0x0}, {0xc000c575d8, 0x2c76aa0})
	/go/src/github.com/influxdata/telegraf/internal/snmp/translate.go:86 +0x38
path/filepath.Walk({0xc000cdf7c0, 0x42}, 0xc000c575d8)
	/usr/local/go/src/path/filepath/path.go:503 +0x50
github.com/influxdata/telegraf/internal/snmp.LoadMibsFromPath({0xc0002722b0, 0x1, 0xc000c576c8}, {0x595c650, 0xc000bda5d0})
	/go/src/github.com/influxdata/telegraf/internal/snmp/translate.go:84 +0x4b7
github.com/influxdata/telegraf/plugins/inputs/snmp.(*Snmp).Init(0xc0003e4f20)
	/go/src/github.com/influxdata/telegraf/plugins/inputs/snmp/snmp.go:103 +0x50
github.com/influxdata/telegraf/models.(*RunningInput).Init(0xc000c57758)
	/go/src/github.com/influxdata/telegraf/models/running_input.go:82 +0x35
github.com/influxdata/telegraf/agent.(*Agent).initPlugins(0xc000482768)
	/go/src/github.com/influxdata/telegraf/agent/agent.go:189 +0x96
github.com/influxdata/telegraf/agent.(*Agent).Run(0xc000482768, {0x58f2a08, 0xc000d0a880})
	/go/src/github.com/influxdata/telegraf/agent/agent.go:105 +0x185
main.runAgent({0x58f2a08, 0xc000d0a880}, {0x84daa60, 0x0, 0x0}, {0x84daa60, 0x0, 0x0})
	/go/src/github.com/influxdata/telegraf/cmd/telegraf/telegraf.go:312 +0xc69
main.reloadLoop({0x84daa60, 0x0, 0x0}, {0x84daa60, 0x0, 0x0})
	/go/src/github.com/influxdata/telegraf/cmd/telegraf/telegraf.go:147 +0x28a
main.run(...)
	/go/src/github.com/influxdata/telegraf/cmd/telegraf/telegraf_posix.go:8
main.main()
	/go/src/github.com/influxdata/telegraf/cmd/telegraf/telegraf.go:485 +0xa9a

System info

telegraf v1.21.3, docker v20.10.3

Docker

Extra mibs are added using apt-get install snmp-mibs-downloader

Dockerfile:

FROM telegraf:1.21.3

# Add deps
RUN sed -i -e's/ main/ main contrib non-free/g' /etc/apt/sources.list && \
    DEBIAN_FRONTEND=noninteractive apt-get update && \
    DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends snmp-mibs-downloader && \
    rm -rf /var/lib/apt/lists/* 
RUN sed -i -e's/mibs /# mibs/g' /etc/snmp/snmp.conf    

Steps to reproduce

1. Build above Dockerfile and use provided telegraf.conf
2. Start telegraf
3. BOOM

Expected behavior

Telegraf should work just as it did in v1.20.3 - i.e. read the extra MIBS and collect the metrics

Actual behavior

Telegraf crash on start

Additional info

No response

[MyaLongmire]

Are you specifying the path where your mibs are stored after you download them? Try adding path = ["/your/path/here"] under version in your config file.

[ptxmac]

Are you specifying the path where your mibs are stored after you download them? Try adding path = ["/your/path/here"] under version in your config file.

That did help me to track the problem down - The crash happens when the mibs path contains a broken symlink. (I run telegraf in docker, and some of the mibs are mounted from the host)

The broken link is a user error, but the previous version of telegraf handled this gracefully instead of crashing

[MyaLongmire]

There is a pr being worked on to check symlinks better and fail more gracefully. Please be patient as we work out all the kinks :)

I'm glad you figured out your issue!

[MyaLongmire]

Sorry I closed this a little fast without more info. Here is the pr that checks symlinks if you want to give it a go or keep an eye out on its progress!