fix: Correct max index while looping over devs

* Ensure we trim spaces in nvidia-smi output

Signed-off-by: Mahendra Paipuri <[email protected]>

pkg/collector/fixtures/nvidia-smi

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-echo """index, name, uuid
+printf """index, name, uuid
 0, Tesla V100-SXM2-32GB, GPU-f124aa59-d406-d45b-9481-8fcd694e6c9e
 1, Tesla V100-SXM2-32GB, GPU-61a65011-6571-a6d2-5ab8-66cbb6f7f9c3
 2, Tesla V100-SXM2-32GB, GPU-61a65011-6571-a6d2-5th8-66cbb6f7f9c3

pkg/collector/helper.go

@@ -118,7 +118,7 @@ func GetNvidiaGPUDevices(nvidiaSmiPath string, logger log.Logger) (map[int]Devic
 	// Get all devices
 	gpuDevices := map[int]Device{}
 	devIndxInt := 0
-	for _, line := range strings.Split(string(nvidiaSmiOutput), "\n") {
+	for _, line := range strings.Split(strings.TrimSpace(string(nvidiaSmiOutput)), "\n") {
 		// Header line, empty line and newlines are ignored
 		if line == "" || line == "\n" || strings.HasPrefix(line, "index") {
 			continue

pkg/collector/slurm.go

@@ -486,7 +486,7 @@ func (c *slurmCollector) getJobProperties(metric *CgroupMetric, pids []uint64) {
 	// it but just to be safe. This will have a small overhead as we need to check the
 	// correct integer index for each device index. We can live with it as there are
 	// typically 2/4/8 GPUs per node.
-	for i := 0; i <= len(c.nvidiaGPUDevs); i++ {
+	for i := 0; i < len(c.nvidiaGPUDevs); i++ {
 		dev := c.nvidiaGPUDevs[i]
 		gpuJobMapInfo := fmt.Sprintf("%s/%s", *gpuStatPath, dev.index)