fix(inputs.infiniband): Handle devices without counters

[jose-d]

• Updated associated README.md.
• Wrote appropriate unit tests.
• Pull request title or commits are in [conventional commit format]

resolves #8135

Not all rdma devices have counters parsable by GetRdmaSysfsStats.
This PR add continue statement to ignore such device (and continue with next device/port) instead of throwing an error and failing

[telegraf-tiger]

Thanks so much for the pull request!
🤝 ✒️ Just a reminder that the CLA has not yet been signed, and we'll need it before merging. Please sign the CLA when you get a chance, then post a comment here saying !signed-cla

[powersj]

Thank you for the PR!

Looking at the issue, it seems that there are devices which may report as infiniband capable, won't have counters, and right now this prevents any metrics. Continuing on allows devices that can correctly get collected report metrics.

Does that sound right?

[jose-d]

!signed-cla

[jose-d]

Looking at the issue, it seems that there are devices which may report as infiniband capable, won't have counters, and right now this prevents any metrics. Continuing on allows devices that can correctly get collected report metrics.

Does that sound right?

Almost :)

inputs.infiniband uses rdmamap library which reports all RDMA-capable (not only Infiniband) devices - and some RDMA devices (eg. ROCE) have no counters, and this stops metric collection. Exactly, as you wrote, continuing on allows devices that can correctly get collected report metrics.

[jose-d]

original behavior (metric collection fails on device irdma1 and no other devices are collected:

[jose@n15 telegraf]$ telegraf --test --config /etc/telegraf/telegraf.d/inputs.infiniband.conf 
...
2023-10-03T21:09:50Z E! [inputs.infiniband] Error in plugin: open /sys/class/infiniband/irdma1/ports/1/counters: no such file or directory
2023-10-03T21:09:50Z E! [telegraf] Error running agent: input plugins recorded 1 errors
[jose@n15 telegraf]$

after suggested patch (metric collection skips irdma1 and sucessfully collects mlx5_0

[jose@n15 telegraf]$ ./telegraf --test --config /etc/telegraf/telegraf.d/inputs.infiniband.conf 
...
> infiniband,device=mlx5_0,host=n15.koios.lan,port=1 VL15_dropped=0i,excessive_buffer_overrun_errors=0i,link_downed=0i,link_error_recovery=0i,local_link_integrity_errors=0i,multicast_rcv_packets=312906i,multicast_xmit_packets=0i,port_rcv_constraint_errors=0i,port_rcv_data=10591292i,port_rcv_errors=0i,port_rcv_packets=318648i,port_rcv_remote_physical_errors=0i,port_rcv_switch_relay_errors=0i,port_xmit_constraint_errors=0i,port_xmit_data=55025i,port_xmit_discards=0i,port_xmit_packets=3736i,port_xmit_wait=0i,symbol_error=0i,unicast_rcv_packets=5742i,unicast_xmit_packets=3736i 1696367342000000000
[jose@n15 telegraf]$

[powersj]

Thanks for the explanation.

It seems the only time errors come from this function are when the file does not exist or we cannot read the file itself, so I think this makes sense that we only return what we can read.

[telegraf-tiger]

Download PR build artifacts for linux_amd64.tar.gz, darwin_amd64.tar.gz, and windows_amd64.zip.
Downloads for additional architectures and packages are available below.

👍 This pull request doesn't change the Telegraf binary size

📦 Click here to get additional PR build artifacts

Artifact URLs

DEB	RPM	TAR GZ	ZIP
amd64.deb	aarch64.rpm	darwin_amd64.tar.gz	windows_amd64.zip
arm64.deb	armel.rpm	darwin_arm64.tar.gz	windows_arm64.zip
armel.deb	armv6hl.rpm	freebsd_amd64.tar.gz	windows_i386.zip
armhf.deb	i386.rpm	freebsd_armv7.tar.gz
i386.deb	ppc64le.rpm	freebsd_i386.tar.gz
mips.deb	riscv64.rpm	linux_amd64.tar.gz
mipsel.deb	s390x.rpm	linux_arm64.tar.gz
ppc64el.deb	x86_64.rpm	linux_armel.tar.gz
riscv64.deb		linux_armhf.tar.gz
s390x.deb		linux_i386.tar.gz
		linux_mips.tar.gz
		linux_mipsel.tar.gz
		linux_ppc64le.tar.gz
		linux_riscv64.tar.gz
		linux_s390x.tar.gz

[srebhan]

Thanks for the fix @jose-d!