Connectrix:Cisco MDS 多個介面丟幀導致服務效能問題
Summary: 多個介面丟幀導致服務效能問題
Symptoms
多個介面丟幀,導致客戶環境出現服務效能問題。
MDS 交換器會記錄下列 PMON 警示:-
記錄記錄:PMON-SLOT1-3-RISING_THRESHOLD_REACHED: Credit Loss Reco has reached the rising threshold PMON-SLOT1-3-RISING_THRESHOLD_REACHED: TX Credit Not Available has reached the rising threshold
Cause
內建記錄記錄審查:
以下 IPA 錯誤表示數據包在收到後正在損壞。 發生這些情況的模組應予以更換`show hardware internal errors`|------------------------------------------------------------------------|| Device:Tbird Que Driver Role:QUE Mod: 1 || Device Statistics Category :: ERROR|------------------------------------------------------------------------|Instance:0Cntr Name Value Ports----- ---- ----- ----- 24 THB_IPA_IPA0_CNT_BAD_CRC 0000000000637620 1-4 -
25 THB_IPA_IPA0_CNT_CORRUPT 0000000000637620 1-4 -
以下 EBM 錯誤與出口線卡收到的損壞數據包相同。這些可能會發生在接收這些損壞數據包的幾個模組上。這些模組沒有故障。更換產生 IPA 錯誤的模組,這些錯誤將停止:`show hardware internal errors`|------------------------------------------------------------------------|| Device:Tbird Fwd Driver Role:L2 Mod: 2 || Device Statistics Category :: ERROR|------------------------------------------------------------------------|Instance:1Cntr Name Value Ports----- ---- ----- ----- 300 THB_EBM0_ACC_ERR_QUEUE_DROP_PKT 0000000000000017 9-12 - <--------
301 THB_EBM0_CNT_ERR_QUEUE_DROP_SF 0000000000000010 9-12 - <--------13414 TBIRD_FWD_EBM_0_PACK0_EPR_DROP 0000000000000001 9-12 - 13422 TBIRD_FWD_EBM_0_PACK1_EPR_DROP 0000000000000005 9-12 - 13430 TBIRD_FWD_EBM_0_PACK2_EPR_DROP 0000000000000001 9-12 -
13438 TBIRD_FWD_EBM_0_PACK3_EPR_DROP 0000000000000003 9-12 - 13727 TBIRD_FWD_EPR0_PKT_CRC_ERR 0000000000000001 9-16 - 13755 TBIRD_FWD_EPR1_PKT_CRC_ERR 0000000000000005 9-16 - 13783 TBIRD_FWD_EPR2_PKT_CRC_ERR 0000000000000001 9-16 -
13811 TBIRD_FWD_EPR3_PKT_CRC_ERR 0000000000000003 9-16 - |------------------------------------------------------------------------|| Device:Tbird Fwd Driver Role:L2 Mod: 3 || Device Statistics Category :: ERROR|------------------------------------------------------------------------|Instance:2Cntr Name Value Ports----- ---- ----- ----- 300 THB_EBM0_ACC_ERR_QUEUE_DROP_PKT 0000000000068016 17-20 -
301 THB_EBM0_CNT_ERR_QUEUE_DROP_SF 0000000000057196 17-20 - |------------------------------------------------------------------------|| Device:Tbird Fwd Driver Role:L2 Mod: 4 || Device Statistics Category :: ERROR|------------------------------------------------------------------------|Instance:0Cntr Name Value Ports----- ---- ----- ----- 300 THB_EBM0_ACC_ERR_QUEUE_DROP_PKT 0000000000004504 1-4 -
301 THB_EBM0_CNT_ERR_QUEUE_DROP_SF 0000000000003598 1-4 - |------------------------------------------------------------------------|| Device:Tbird Fwd Driver Role:L2 Mod: 5 || Device Statistics Category :: ERROR|------------------------------------------------------------------------| 300 THB_EBM0_ACC_ERR_QUEUE_DROP_PKT 0000000000020069 1-4 -
301 THB_EBM0_CNT_ERR_QUEUE_DROP_SF 0000000000019715 1-4 -
裸金屬裸金屬伺服器錯誤與出口線卡收到的損壞數據包相同。這些可能會發生在接收這些損壞數據包的幾個模組上。這些模組沒有故障。更換產生 IPA 錯誤的模組,這些錯誤將會停止:
僅發生在 MDS DS-X9232-256K9 和 DS-X9248-256K9 模組上。
Resolution
定:
更換記錄下列錯誤的線路卡:THB_IPA_IPA0_CNT_BAD_CRC THB_IPA_IPA0_CNT_CORRUPT
注意:埠監視器有 3 個特殊計數器,可針對上述情況發出警報。由於預設的「slowdrain」策略大多使用,因此這些策略不被監視。
計數器是:
err-pkt-from-port err-pkt-to-xbar
err-pkt-from-xbar
另請參閱其他資訊
:CSCum09652 新增 Gen4 IPA 和 EFI 錯誤至連接埠監控 xbar 錯誤