mchus/logpile
1
0
Fork 0
Code Issues 5 Pull Requests Actions Packages Projects Releases 9 Wiki Activity

fix: dedup GPUs across multiple chassis PCIeDevice trees in Redfish collector

Browse Source 
Supermicro HGX exposes each GPU under both Chassis/1/PCIeDevices and a
dedicated Chassis/HGX_GPU_SXM_N/PCIeDevices. gpuDocDedupKey was keying
by @odata.id path, so identical GPUs with the same serial were not
deduplicated across sources. Now stable identifiers (serial → BDF →
slot+model) take priority over path.

Also includes Inspur parser improvements: NVMe model/serial enrichment
from devicefrusdr.log and audit.log, RAID drive slot normalization to
BP notation, PSU slot normalization, BMC/CPLD/VR firmware from RESTful
version info section, and parser version bump to 1.8.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Mikhail Chusavitin
2026-03-06 14:44:36 +03:00
parent 62d6ad6f66
commit 9df29b1be9
9 changed files with 362 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
internal/parser/vendors/inspur/parser.go vendored
View File

@@ -16,7 +16,7 @@ import (
// parserVersion - version of this parser module
// IMPORTANT: Increment this version when making changes to parser logic!
const parserVersion = "1.5"
const parserVersion = "1.8"
func init() {
parser.Register(&Parser{})
@@ -95,9 +95,41 @@ func (p *Parser) Parse(files []parser.ExtractedFile) (*models.AnalysisResult, er
Sensors: make([]models.SensorReading, 0),
}
// Pre-parse enrichment maps from devicefrusdr.log for use inside ParseAssetJSON.
// BMC does not populate HddInfo.ModelName or SerialNumber for NVMe drives.
var pcieSlotDeviceNames map[int]string
var nvmeLocToSlot map[int]int
if f := parser.FindFileByName(files, "devicefrusdr.log"); f != nil {
pcieSlotDeviceNames = ParsePCIeSlotDeviceNames(f.Content)
nvmeLocToSlot = ParsePCIeNVMeLocToSlot(f.Content)
}
// Parse NVMe serial numbers from audit.log: every disk SN change is logged there.
// Combine with the NVMe loc→slot mapping to build pcieSlot→serial map.
// Also parse RAID disk serials by backplane slot key (e.g. "BP0:0").
var pcieSlotSerials map[int]string
var raidSlotSerials map[string]string
if f := parser.FindFileByName(files, "audit.log"); f != nil {
if len(nvmeLocToSlot) > 0 {
nvmeDiskSerials := ParseAuditLogNVMeSerials(f.Content)
if len(nvmeDiskSerials) > 0 {
pcieSlotSerials = make(map[int]string, len(nvmeDiskSerials))
for diskNum, serial := range nvmeDiskSerials {
if slot, ok := nvmeLocToSlot[diskNum]; ok {
pcieSlotSerials[slot] = serial
}
}
if len(pcieSlotSerials) == 0 {
pcieSlotSerials = nil
}
}
}
raidSlotSerials = ParseAuditLogRAIDSerials(f.Content)
}
// Parse asset.json first (base hardware info)
if f := parser.FindFileByName(files, "asset.json"); f != nil {
if hw, err := ParseAssetJSON(f.Content); err == nil {
if hw, err := ParseAssetJSON(f.Content, pcieSlotDeviceNames, pcieSlotSerials); err == nil {
result.Hardware = hw
}
}
@@ -182,6 +214,9 @@ func (p *Parser) Parse(files []parser.ExtractedFile) (*models.AnalysisResult, er
if result.Hardware != nil {
applyGPUStatusFromEvents(result.Hardware, result.Events)
enrichStorageFromSerialFallbackFiles(files, result.Hardware)
// Apply RAID disk serials from audit.log (authoritative: last non-NULL SN change).
// These override redis/component.log serials which may be stale after disk replacement.
applyRAIDSlotSerials(result.Hardware, raidSlotSerials)
}
return result, nil