Fixes #20. This onekeylog variant has no devicefrusdr.log at all: FRU/sensors come from raw ipmitool text output, PCIe/GPU presence has a dedicated structural snapshot, SEL lives at a different path, and BMC component failures are logged separately from SEL/IDL. - Fall back to component/fru.txt (same FRU block format as devicefrusdr.log) and component/sensor.txt / sdr.txt (ipmitool sensor list / sdr elist) when devicefrusdr.log is absent. - Parse log/bmc/diagnose/OtrdDiagnoseComponent.json's PCIe Device Info array for GPU/PCIe presence and link state, independent of SEL/IDL alarm history; flag devices running below their negotiated max link speed/width as degraded with a Warning event. - Fall back to log/sel.csv (same format as selelist.csv) when selelist.csv is absent. - Parse log/bmc/commer-comp/{commerslot,commerhmc,commerswvr,commerswcpld} logs (including rotated *.tar.gz.N parts) into failure events, filtering known-noisy lines. - Collapse SEL events duplicated across sources by (timestamp, event_type, description). - Surface a CollectionError when FRU/sensors are still empty after all fallbacks, instead of silently returning an empty inventory. - Fix ParseFRU: a later placeholder "Product Serial : 0" / "Product Part Number : NULL" line in the same FRU block (e.g. SCM_FRU) was overwriting an already-parsed real Board Serial/Part Number. Verified against dump_23DB01633_20260727-1359.tar.gz (HGX B200, KR9288-X3): fru 0→21, sensors 0→303, 8 GPUs present at Gen5 x16 in slots 100-107. Deferred (not covered by this change): BIOS-change-settings context and BIOS POST codes from the same layout — see bible-local/10-decisions.md ADL-048. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
168 lines
4.8 KiB
Go
168 lines
4.8 KiB
Go
package inspur
|
|
|
|
import (
|
|
"bufio"
|
|
"regexp"
|
|
"strings"
|
|
|
|
"git.mchus.pro/mchus/logpile/internal/models"
|
|
)
|
|
|
|
var (
|
|
fruDeviceRegex = regexp.MustCompile(`^FRU Device Description\s*:\s*(.+)$`)
|
|
fruFieldRegex = regexp.MustCompile(`^\s+(.+?)\s*:\s*(.*)$`)
|
|
platformIdRegex = regexp.MustCompile(`(?i)PlatformId\s*=\s*(\S+)`)
|
|
)
|
|
|
|
// ParseFRU parses BMC FRU (Field Replaceable Unit) output
|
|
func ParseFRU(content []byte) []models.FRUInfo {
|
|
var fruList []models.FRUInfo
|
|
var current *models.FRUInfo
|
|
|
|
scanner := bufio.NewScanner(strings.NewReader(string(content)))
|
|
for scanner.Scan() {
|
|
line := scanner.Text()
|
|
|
|
// Check for new FRU device
|
|
if matches := fruDeviceRegex.FindStringSubmatch(line); matches != nil {
|
|
if current != nil && current.Description != "" {
|
|
fruList = append(fruList, *current)
|
|
}
|
|
current = &models.FRUInfo{
|
|
Description: strings.TrimSpace(matches[1]),
|
|
}
|
|
continue
|
|
}
|
|
|
|
// Skip if no current FRU device
|
|
if current == nil {
|
|
continue
|
|
}
|
|
|
|
// Skip "Device not present" entries
|
|
if strings.Contains(line, "Device not present") {
|
|
current = nil
|
|
continue
|
|
}
|
|
|
|
// Parse FRU fields
|
|
if matches := fruFieldRegex.FindStringSubmatch(line); matches != nil {
|
|
fieldName := strings.TrimSpace(matches[1])
|
|
fieldValue := strings.TrimSpace(matches[2])
|
|
|
|
switch fieldName {
|
|
case "Chassis Type":
|
|
current.ChassisType = fieldValue
|
|
case "Chassis Part Number":
|
|
if fieldValue != "0" {
|
|
current.PartNumber = fieldValue
|
|
}
|
|
case "Chassis Serial":
|
|
if fieldValue != "0" {
|
|
current.SerialNumber = fieldValue
|
|
}
|
|
case "Board Mfg Date":
|
|
current.MfgDate = fieldValue
|
|
case "Board Mfg", "Product Manufacturer":
|
|
if fieldValue != "NULL" {
|
|
current.Manufacturer = fieldValue
|
|
}
|
|
case "Board Product", "Product Name":
|
|
if fieldValue != "NULL" {
|
|
current.ProductName = fieldValue
|
|
}
|
|
case "Board Serial", "Product Serial":
|
|
// Modules with no product-level FRU (e.g. SCM_FRU) report
|
|
// "Product Serial : 0" after a real "Board Serial" value;
|
|
// don't let the placeholder overwrite the real serial.
|
|
if fieldValue != "0" && fieldValue != "NULL" {
|
|
current.SerialNumber = fieldValue
|
|
}
|
|
case "Board Part Number", "Product Part Number":
|
|
// See "Board Serial" above: placeholder "NULL"/"0" product-level
|
|
// fields must not overwrite a real board-level part number.
|
|
if fieldValue != "0" && fieldValue != "NULL" {
|
|
current.PartNumber = fieldValue
|
|
}
|
|
case "Product Version":
|
|
if fieldValue != "0" {
|
|
current.Version = fieldValue
|
|
}
|
|
case "Product Asset Tag":
|
|
if fieldValue != "NULL" {
|
|
current.AssetTag = fieldValue
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Don't forget the last one
|
|
if current != nil && current.Description != "" {
|
|
fruList = append(fruList, *current)
|
|
}
|
|
|
|
return fruList
|
|
}
|
|
|
|
// extractBoardInfo extracts main board/chassis information from FRU data
|
|
func extractBoardInfo(fruList []models.FRUInfo, hw *models.HardwareConfig) {
|
|
if hw == nil || len(fruList) == 0 {
|
|
return
|
|
}
|
|
|
|
// Look for the main board/chassis FRU entry.
|
|
// Keep the first non-empty serial as the server serial and avoid overwriting it
|
|
// with module-specific serials (e.g., SCM_FRU).
|
|
for _, fru := range fruList {
|
|
// Skip empty entries
|
|
if fru.ProductName == "" && fru.SerialNumber == "" {
|
|
continue
|
|
}
|
|
|
|
// Prioritize entries that look like main board info
|
|
desc := strings.ToLower(fru.Description)
|
|
isMainBoard := strings.Contains(desc, "builtin") ||
|
|
strings.Contains(desc, "fru device") ||
|
|
strings.Contains(desc, "chassis") ||
|
|
strings.Contains(desc, "board")
|
|
|
|
if fru.SerialNumber != "" && hw.BoardInfo.SerialNumber == "" {
|
|
hw.BoardInfo.SerialNumber = fru.SerialNumber
|
|
}
|
|
if fru.ProductName != "" && (hw.BoardInfo.ProductName == "" || isMainBoard) {
|
|
hw.BoardInfo.ProductName = fru.ProductName
|
|
}
|
|
// Manufacturer from non-main FRU entries (e.g. PSU vendor) should not become server vendor.
|
|
if fru.Manufacturer != "" && isMainBoard && hw.BoardInfo.Manufacturer == "" {
|
|
hw.BoardInfo.Manufacturer = fru.Manufacturer
|
|
}
|
|
if fru.PartNumber != "" && (hw.BoardInfo.PartNumber == "" || isMainBoard) {
|
|
hw.BoardInfo.PartNumber = fru.PartNumber
|
|
}
|
|
|
|
// Main board entry with complete data is good enough to stop.
|
|
if isMainBoard && hw.BoardInfo.ProductName != "" && hw.BoardInfo.SerialNumber != "" {
|
|
break
|
|
}
|
|
}
|
|
}
|
|
|
|
// extractPlatformId extracts server model from ThermalConfig (PlatformId)
|
|
func extractPlatformId(content []byte, hw *models.HardwareConfig) {
|
|
if hw == nil {
|
|
return
|
|
}
|
|
|
|
if match := platformIdRegex.FindSubmatch(content); match != nil {
|
|
platformId := strings.TrimSpace(string(match[1]))
|
|
if platformId != "" {
|
|
// Set as ProductName (server model) - this takes priority over FRU data
|
|
hw.BoardInfo.ProductName = platformId
|
|
// Also set manufacturer as Inspur if not already set
|
|
if hw.BoardInfo.Manufacturer == "" {
|
|
hw.BoardInfo.Manufacturer = "Inspur"
|
|
}
|
|
}
|
|
}
|
|
}
|