long node name causes index out of range error

[hokiegeek2]

First of all, great project! This works great.

Found one easily fixable issue. If the node names are too long, the memoryAlloc field value merges into the node name value:

INFO[0000] Starting Server: 0.0.0.0:9090                 source="main.go:59"
INFO[0000] GPUs Accounting: false                        source="main.go:60"
INFO[0049] sinfo fields: [production-slurm-compute-10 8192 0/4/0/4 idle]  source="node.go:55"
panic: runtime error: index out of range [4] with length 4

goroutine 39 [running]:
main.ParseNodeMetrics({0xc0001b9000, 0x594, 0xc0000adde0?})
	/home/jyost/development/git/prometheus-slurm-exporter/node.go:58 +0x5a5
main.NodeGetMetrics()
	/home/jyost/development/git/prometheus-slurm-exporter/node.go:41 +0x1e
main.(*NodeCollector).Collect(0xc00007f740, 0xc0000adf60?)
	/home/jyost/development/git/prometheus-slurm-exporter/node.go:131 +0x3e
github.com/prometheus/client_golang/prometheus.(*Registry).Gather.func1()
	/home/jyost/development/git/prometheus-slurm-exporter/go/modules/pkg/mod/github.com/prometheus/[email protected]/prometheus/registry.go:443 +0xfb
created by github.com/prometheus/client_golang/prometheus.(*Registry).Gather
	/home/jyost/development/git/prometheus-slurm-exporter/go/modules/pkg/mod/github.com/prometheus/[email protected]/prometheus/registry.go:535 +0xb0b

The node name is production-slurm-compute-1 and the 0 got appended to the end of the node name. I fixed in my fork and will submit a PR soon

[hokiegeek2]

whoops, this issue is a dupe