feat: HPE iLO support — profile, post-probe hang fix, replay parser fixes, AHS parser

- Add HPE iLO Redfish profile (priority 20): matches on manufacturer/OEM/iLO signals,
  adds SmartStorage/SmartStorageConfig to critical paths, sets realistic ETA baseline
  and rate policy for iLO's known slowness
- Fix post-probe hang on HPE iLO: skip numeric probing of collections where
  Members@odata.count == len(Members); add 4s postProbeClient timeout as safety net
- Exclude /WorkloadPerformanceAdvisor from crawl paths
- Fix replay parser: skip absent CPU sockets, absent DIMM slots, absent drive bays
- Filter N/A version entries from firmware inventory
- Remove drive firmware from general firmware list (already in Storage[].Firmware)
- Add HPE AHS (.ahs) archive parser with hybrid SMBIOS/Redfish extraction

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Mikhail Chusavitin
2026-03-30 13:39:14 +03:00
parent 3547ef9083
commit 0f310d57c4
14 changed files with 1729 additions and 12 deletions
+32 -1
View File
@@ -1606,6 +1606,7 @@ func (c *RedfishConnector) collectRawRedfishTree(ctx context.Context, client *ht
crawlStart := time.Now()
memoryClient := c.httpClientWithTimeout(req, redfishSnapshotMemoryRequestTimeout())
memoryGate := make(chan struct{}, redfishSnapshotMemoryConcurrency())
postProbeClient := c.httpClientWithTimeout(req, redfishSnapshotPostProbeRequestTimeout())
branchLimiter := newRedfishSnapshotBranchLimiter(redfishSnapshotBranchConcurrency())
branchRetryPause := redfishSnapshotBranchRequeueBackoff()
timings := newRedfishPathTimingCollector(4)
@@ -1908,7 +1909,7 @@ func (c *RedfishConnector) collectRawRedfishTree(ctx context.Context, client *ht
ETASeconds: int(estimateProgressETA(postProbeStart, i, len(postProbeCollections), 3*time.Second).Seconds()),
})
}
for childPath, doc := range c.probeDirectRedfishCollectionChildren(ctx, client, req, baseURL, path) {
for childPath, doc := range c.probeDirectRedfishCollectionChildren(ctx, postProbeClient, req, baseURL, path) {
if _, exists := out[childPath]; exists {
continue
}
@@ -2158,6 +2159,12 @@ func shouldAdaptivePostProbeCollectionPath(path string, collectionDoc map[string
if len(memberRefs) == 0 {
return true
}
// If the collection reports an explicit non-zero member count that already
// matches the number of discovered member refs, every member is accounted
// for and numeric probing cannot find anything new.
if odataCount := asInt(collectionDoc["Members@odata.count"]); odataCount > 0 && odataCount == len(memberRefs) {
return false
}
return redfishCollectionHasNumericMemberRefs(memberRefs)
}
@@ -2350,6 +2357,18 @@ func redfishSnapshotRequestTimeout() time.Duration {
return 12 * time.Second
}
func redfishSnapshotPostProbeRequestTimeout() time.Duration {
if v := strings.TrimSpace(os.Getenv("LOGPILE_REDFISH_POSTPROBE_TIMEOUT")); v != "" {
if d, err := time.ParseDuration(v); err == nil && d > 0 {
return d
}
}
// Post-probe probes non-existent numeric paths expecting fast 404s.
// A short timeout prevents BMCs that hang on unknown paths from stalling
// the entire collection for minutes (e.g. HPE iLO on NetworkAdapters Ports).
return 4 * time.Second
}
func redfishSnapshotWorkers(tuning redfishprofile.AcquisitionTuning) int {
if tuning.SnapshotWorkers >= 1 && tuning.SnapshotWorkers <= 16 {
return tuning.SnapshotWorkers
@@ -2852,6 +2871,8 @@ func shouldCrawlPath(path string) bool {
"/GetServerAllUSBStatus",
"/Oem/Public/KVM",
"/SecureBoot/SecureBootDatabases",
// HPE iLO WorkloadPerformanceAdvisor — operational/advisory data, not inventory.
"/WorkloadPerformanceAdvisor",
} {
if strings.Contains(normalized, part) {
return false
@@ -5047,6 +5068,16 @@ func isVirtualStorageDrive(doc map[string]interface{}) bool {
return false
}
// isAbsentDriveDoc returns true when the drive document represents an empty bay
// with no installed media (Status.State == "Absent"). These should be excluded
// from the storage inventory.
func isAbsentDriveDoc(doc map[string]interface{}) bool {
if status, ok := doc["Status"].(map[string]interface{}); ok {
return strings.EqualFold(asString(status["State"]), "Absent")
}
return strings.EqualFold(asString(doc["Status"]), "Absent")
}
func looksLikeDrive(doc map[string]interface{}) bool {
if asString(doc["MediaType"]) != "" {
return true
+41
View File
@@ -96,6 +96,7 @@ func ReplayRedfishFromRawPayloads(rawPayloads map[string]any, emit ProgressFn) (
networkProtocolDoc, _ := r.getJSON(joinPath(primaryManager, "/NetworkProtocol"))
firmware := parseFirmware(systemDoc, biosDoc, managerDoc, networkProtocolDoc)
firmware = dedupeFirmwareInfo(append(firmware, r.collectFirmwareInventory()...))
firmware = filterStorageDriveFirmware(firmware, storageDevices)
bmcManagementSummary := r.collectBMCManagementSummary(managerPaths)
boardInfo.BMCMACAddress = strings.TrimSpace(firstNonEmpty(
asString(bmcManagementSummary["mac_address"]),
@@ -498,6 +499,10 @@ func (r redfishSnapshotReader) collectFirmwareInventory() []models.FirmwareInfo
if strings.TrimSpace(version) == "" {
continue
}
// Skip placeholder version strings that carry no useful information.
if strings.EqualFold(strings.TrimSpace(version), "N/A") {
continue
}
name := firmwareInventoryDeviceName(doc)
name = strings.TrimSpace(name)
if name == "" {
@@ -550,6 +555,32 @@ func dedupeFirmwareInfo(items []models.FirmwareInfo) []models.FirmwareInfo {
return out
}
// filterStorageDriveFirmware removes from fw any entries whose DeviceName+Version
// already appear as a storage drive's Model+Firmware. Drive firmware is already
// represented in the Storage section and should not be duplicated in the general
// firmware list.
func filterStorageDriveFirmware(fw []models.FirmwareInfo, storage []models.Storage) []models.FirmwareInfo {
if len(storage) == 0 {
return fw
}
driveFW := make(map[string]struct{}, len(storage))
for _, d := range storage {
model := strings.ToLower(strings.TrimSpace(d.Model))
rev := strings.ToLower(strings.TrimSpace(d.Firmware))
if model != "" && rev != "" {
driveFW[model+"|"+rev] = struct{}{}
}
}
out := fw[:0:0]
for _, f := range fw {
key := strings.ToLower(strings.TrimSpace(f.DeviceName)) + "|" + strings.ToLower(strings.TrimSpace(f.Version))
if _, skip := driveFW[key]; !skip {
out = append(out, f)
}
}
return out
}
func (r redfishSnapshotReader) collectThresholdSensors(chassisPaths []string) []models.SensorReading {
out := make([]models.SensorReading, 0)
seen := make(map[string]struct{})
@@ -1261,6 +1292,12 @@ func (r redfishSnapshotReader) collectProcessors(systemPath string) []models.CPU
!strings.EqualFold(pt, "CPU") && !strings.EqualFold(pt, "General") {
continue
}
// Skip absent processor sockets — empty slots with no CPU installed.
if status, ok := doc["Status"].(map[string]interface{}); ok {
if strings.EqualFold(asString(status["State"]), "Absent") {
continue
}
}
cpu := parseCPUs([]map[string]interface{}{doc})[0]
if cpu.Socket == 0 && socketIdx > 0 && strings.TrimSpace(asString(doc["Socket"])) == "" {
cpu.Socket = socketIdx
@@ -1287,6 +1324,10 @@ func (r redfishSnapshotReader) collectMemory(systemPath string) []models.MemoryD
out := make([]models.MemoryDIMM, 0, len(memberDocs))
for _, doc := range memberDocs {
dimm := parseMemory([]map[string]interface{}{doc})[0]
// Skip empty DIMM slots — no installed memory.
if !dimm.Present {
continue
}
supplementalDocs := r.getLinkedSupplementalDocs(doc, "MemoryMetrics", "EnvironmentMetrics", "Metrics")
if len(supplementalDocs) > 0 {
dimm.Details = mergeGenericDetails(dimm.Details, redfishMemoryDetailsAcrossDocs(doc, supplementalDocs...))
+12 -9
View File
@@ -14,13 +14,16 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
driveDocs, err := r.getCollectionMembers(driveCollectionPath)
if err == nil {
for _, driveDoc := range driveDocs {
if !isVirtualStorageDrive(driveDoc) {
if !isAbsentDriveDoc(driveDoc) && !isVirtualStorageDrive(driveDoc) {
supplementalDocs := r.getLinkedSupplementalDocs(driveDoc, "DriveMetrics", "EnvironmentMetrics", "Metrics")
out = append(out, parseDriveWithSupplementalDocs(driveDoc, supplementalDocs...))
}
}
if len(driveDocs) == 0 {
for _, driveDoc := range r.probeDirectDiskBayChildren(driveCollectionPath) {
if isAbsentDriveDoc(driveDoc) {
continue
}
supplementalDocs := r.getLinkedSupplementalDocs(driveDoc, "DriveMetrics", "EnvironmentMetrics", "Metrics")
out = append(out, parseDriveWithSupplementalDocs(driveDoc, supplementalDocs...))
}
@@ -43,7 +46,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
if err != nil {
continue
}
if !isVirtualStorageDrive(driveDoc) {
if !isAbsentDriveDoc(driveDoc) && !isVirtualStorageDrive(driveDoc) {
supplementalDocs := r.getLinkedSupplementalDocs(driveDoc, "DriveMetrics", "EnvironmentMetrics", "Metrics")
out = append(out, parseDriveWithSupplementalDocs(driveDoc, supplementalDocs...))
}
@@ -51,7 +54,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
continue
}
if looksLikeDrive(member) {
if isVirtualStorageDrive(member) {
if isAbsentDriveDoc(member) || isVirtualStorageDrive(member) {
continue
}
supplementalDocs := r.getLinkedSupplementalDocs(member, "DriveMetrics", "EnvironmentMetrics", "Metrics")
@@ -63,14 +66,14 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
driveDocs, err := r.getCollectionMembers(joinPath(enclosurePath, "/Drives"))
if err == nil {
for _, driveDoc := range driveDocs {
if looksLikeDrive(driveDoc) && !isVirtualStorageDrive(driveDoc) {
if looksLikeDrive(driveDoc) && !isAbsentDriveDoc(driveDoc) && !isVirtualStorageDrive(driveDoc) {
supplementalDocs := r.getLinkedSupplementalDocs(driveDoc, "DriveMetrics", "EnvironmentMetrics", "Metrics")
out = append(out, parseDriveWithSupplementalDocs(driveDoc, supplementalDocs...))
}
}
if len(driveDocs) == 0 {
for _, driveDoc := range r.probeDirectDiskBayChildren(joinPath(enclosurePath, "/Drives")) {
if isVirtualStorageDrive(driveDoc) {
if isAbsentDriveDoc(driveDoc) || isVirtualStorageDrive(driveDoc) {
continue
}
out = append(out, parseDrive(driveDoc))
@@ -83,7 +86,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
if len(plan.KnownStorageDriveCollections) > 0 {
for _, driveDoc := range r.collectKnownStorageMembers(systemPath, plan.KnownStorageDriveCollections) {
if looksLikeDrive(driveDoc) && !isVirtualStorageDrive(driveDoc) {
if looksLikeDrive(driveDoc) && !isAbsentDriveDoc(driveDoc) && !isVirtualStorageDrive(driveDoc) {
supplementalDocs := r.getLinkedSupplementalDocs(driveDoc, "DriveMetrics", "EnvironmentMetrics", "Metrics")
out = append(out, parseDriveWithSupplementalDocs(driveDoc, supplementalDocs...))
}
@@ -98,7 +101,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
}
for _, devAny := range devices {
devDoc, ok := devAny.(map[string]interface{})
if !ok || !looksLikeDrive(devDoc) || isVirtualStorageDrive(devDoc) {
if !ok || !looksLikeDrive(devDoc) || isAbsentDriveDoc(devDoc) || isVirtualStorageDrive(devDoc) {
continue
}
out = append(out, parseDrive(devDoc))
@@ -112,7 +115,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
continue
}
for _, driveDoc := range driveDocs {
if !looksLikeDrive(driveDoc) || isVirtualStorageDrive(driveDoc) {
if !looksLikeDrive(driveDoc) || isAbsentDriveDoc(driveDoc) || isVirtualStorageDrive(driveDoc) {
continue
}
out = append(out, parseDrive(driveDoc))
@@ -124,7 +127,7 @@ func (r redfishSnapshotReader) collectStorage(systemPath string, plan redfishpro
continue
}
for _, driveDoc := range r.probeSupermicroNVMeDiskBays(chassisPath) {
if !looksLikeDrive(driveDoc) || isVirtualStorageDrive(driveDoc) {
if !looksLikeDrive(driveDoc) || isAbsentDriveDoc(driveDoc) || isVirtualStorageDrive(driveDoc) {
continue
}
out = append(out, parseDrive(driveDoc))
@@ -0,0 +1,67 @@
package redfishprofile
func hpeProfile() Profile {
return staticProfile{
name: "hpe",
priority: 20,
safeForFallback: true,
matchFn: func(s MatchSignals) int {
score := 0
if containsFold(s.SystemManufacturer, "hpe") ||
containsFold(s.SystemManufacturer, "hewlett packard") ||
containsFold(s.ChassisManufacturer, "hpe") ||
containsFold(s.ChassisManufacturer, "hewlett packard") {
score += 80
}
for _, ns := range s.OEMNamespaces {
if containsFold(ns, "hpe") {
score += 30
break
}
}
if containsFold(s.ServiceRootProduct, "ilo") {
score += 30
}
if containsFold(s.ManagerManufacturer, "hpe") || containsFold(s.ManagerManufacturer, "ilo") {
score += 20
}
return min(score, 100)
},
extendAcquisition: func(plan *AcquisitionPlan, _ MatchSignals) {
// HPE ProLiant SmartStorage RAID controller inventory is not reachable
// via standard Redfish Storage paths — it requires the HPE OEM SmartStorage tree.
ensureScopedPathPolicy(plan, AcquisitionScopedPathPolicy{
SystemCriticalSuffixes: []string{
"/SmartStorage",
"/SmartStorageConfig",
},
ManagerCriticalSuffixes: []string{
"/LicenseService",
},
})
// HPE iLO responds more slowly than average BMCs under load; give the
// ETA estimator a realistic baseline so progress reports are accurate.
ensureETABaseline(plan, AcquisitionETABaseline{
DiscoverySeconds: 12,
SnapshotSeconds: 180,
PrefetchSeconds: 30,
CriticalPlanBSeconds: 40,
ProfilePlanBSeconds: 25,
})
ensureRecoveryPolicy(plan, AcquisitionRecoveryPolicy{
EnableProfilePlanB: true,
})
// HPE iLO starts throttling under high request rates. Setting a higher
// latency tolerance prevents the adaptive throttler from treating normal
// iLO slowness as a reason to stall the collection.
ensureRatePolicy(plan, AcquisitionRatePolicy{
TargetP95LatencyMS: 1200,
ThrottleP95LatencyMS: 2500,
MinSnapshotWorkers: 2,
MinPrefetchWorkers: 1,
DisablePrefetchOnErrors: true,
})
addPlanNote(plan, "hpe ilo acquisition extensions enabled")
},
}
}
@@ -55,6 +55,7 @@ func BuiltinProfiles() []Profile {
msiProfile(),
supermicroProfile(),
dellProfile(),
hpeProfile(),
inspurGroupOEMPlatformsProfile(),
hgxProfile(),
xfusionProfile(),