Troubleshoot NCS6K Card Boot Failures
Available Languages
Bias-Free Language
The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.
Contents
Introduction
This document describes how to troubleshoot Network Convergence System 6000 (NCS6K) Line card boot failures. In addition, it also provides an overview of data that can be collected which can help TAC to conduct a detailed investigation.
Prerequisites
Requirements
Cisco recommends that you have basic knowledge of XR Command Line Interface (CLI).
Components Used
This document was created using XR release versions 5.0.1, 5.2.1, 5.2.3 and 5.2.4.
The information in this document was created from the devices in a specific lab environment. All of the devices used in this document started with a cleared (default) configuration. If your network is live, make sure that you understand the potential impact of any command.
Background Information
If Route Processor (RP), Fabric Card (FC) or Line Card (LC) is not booting and getting stuck in the boot stage, the first step should be to start collecting show tech ctrace from Syadmin Virtual Machine (VM). This show tech provides information on the current system state and interaction between different components. However, there is a chance that Sysadmin also known as Calvados is not up on the card and show tech ctrace will not be able to collect information for the affected card. This will occur because the card is not reachable via Secure Shell (SSH). For such cases, rconsole procedure will be necessary in order to know why the card gets stuck at the time of boot process.
Note: This file is usually pretty big (500MB-1GB) and will be stored on the Sysadmin VM. In order to extract it from the box, it must be copied to the XR VM (instructions provided later in the same document).
Troubleshoot
Check the status of each card on the Sysadmin VM and find out its current state. Pay special attention to H/w and S/w State. Note that cards that show S/w state as N/A are CPU less cards (FC cards, Fan Tray, etc.) which are essentially controlled by RP CPU. Cards in operational state are CPU based cards and therefore have the software loaded.
sysadmin-vm:0_RP0# show platform
Location Card Type HW State SW State Config State
----------------------------------------------------------------------------
0/0 PROTO-CXP-1XPITA OPERATIONAL OPERATIONAL NSHUT
0/2 PROTO-CXP-2XPITA POWERED_OFF SW_INACTIVE NSHUT
0/3 NC6-10X100G-M-K OPERATIONAL OPERATIONAL NSHUT
0/RP0 NC6-RP OPERATIONAL OPERATIONAL NSHUT
0/RP1 NC6-RP POWERED_ON SW_INACTIVE NSHUT
0/FC0 NC6-FC POWERED_ON N/A NSHUT
0/FC1 NC6-FC-MC POWERED_ON N/A NSHUT
0/FC2 UNKNOWN FAILED N/A NSHUT
0/FC3 NC6-FC POWERED_ON N/A NSHUT
0/FC4 NC6-FC-B2B POWERED_ON N/A NSHUT
0/FC5 NC6-FC OPERATIONAL N/A NSHUT
0/FT0 NC6-FANTRAY OPERATIONAL N/A NSHUT
0/FT1 NC6-FANTRAY OPERATIONAL N/A NSHUT
0/PT0 NCS-AC-PWRTRAY OPERATIONAL N/A NSHUT
0/PT1 NCS-AC-PWRTRAY OPERATIONAL N/A NSHUT
0/PT2 NCS-AC-PWRTRAY OPERATIONAL N/A NSHUT
Next step is to check Card Chip Controller (CCC) inventory command and confirm the card state.
sysadmin-vm:0_RP0# show controller ccc inventory summary
CCC Inventory Summary :
BP HW
Location Card Type ID Serial Number Ver Card State
--------------------------------------------------------------------------------
0/RP0 NC6-RP (master) 0 SAD15270129 0.1 CARD_READY
0/RP1 NC6-RP (slave) 1 SAD1527012P 0.1 CARD_READY
0/FC0 NC6-FC 8 SAD1618002F 0.2 WAIT_DEV_INIT
0/FC1 NC6-FC 9 SAD153901ZT 0.2 WAIT_DEV_INIT
0/FC4 NC6-FC 12 SAL1803KQEY 1.0 PON_POWERING_UP
0/FC5 NC6-FC 13 SAD16180043 0.2 WAIT_DEV_INIT
0/0 NC6-10X100G-M-K 16 SAL1650UCN9 0.4 PXE_BOOTING
0/4 NC6-10X100G-M-K 20 SAD154502XU 0.1 CARD_READY
Here are the different scenarios that can take place. Also, listed here are expected outputs and next troubleshooting steps.
Linecard Booted as Expected
sysadmin-vm:F0_SC0# show platform detail location 0/0
Platform Information for 0/0
PID : NC6-10X100G-M-P
Description : "NCS 6000 10x100G Multi-Service CXP"
VID/SN : V01
HW Oper State : OPERATIONAL
SW Oper State : OPERATIONAL
Configuration : "NSHUT RST"
HW Version : 1.0
Last Event : HW_EVENT_OK
Last Event Reason : "Initial discovered state:BOOTED (card ok)"
The output of Last Event and Last Event Reason show that card is okay. Execute show reboot history command in order to validate if this specific card had issues in the past and if yes, what was the issue.
sysadmin-vm:F0_SC0# show reboot-history card location 0/0
Card Reboot History for 0/0
0
Timestamp "Fri Oct 2 15:15:26 2015"
Reason Code 7
Reason "Install Activate System Reload"
Src Location ""
Src Name INSTALL
Timestamp "Tue Sep 8 18:56:29 2015"
Reason Code 7
Reason "ADMIN CLI RELOAD ROUTER GRACEFUL"
Src Location ""
Src Name "CONFD USER"
Aborted: by user
Linecard 0/0 is operational and the reason for the last reload was install which essentially means either Software Maintenance Update (SMU) installation or software upgrade was done. This is expected and thus there were no issues with this card.
Fabric Card is Unable to Boot due to Driver Error
sysadmin-vm:0_RP0# show platform
Location Card Type HW State SW State Config State
----------------------------------------------------------------------------
0/FC2 UNKNOWN FAILED N/A NSHUT <-
sysadmin-vm:F0_SC0# show reboot-history card location 0/FC2
Card Reboot History for 0/FC2
Timestamp "Thu Oct 9 12:10:22 2014"
Reason Code 15 <--
"Board reload as devices not up on Fabric Card"
Src Location 0/FC0
Src Name FAM_AGENT_CALV_DRIVER_SFE ß
In this example, FC card did not boot because the sfe_driver was not initialized properly.
Execute few more commands in order to view the reset-history from the CCC point of view. The reboot-history CLI can be used in conjunction with CCC Driver's reset-history CLI to determine the source and reason of card reload.
There could be two options:
Onboard - Use this option to look for information if the card went through Warm Reset (just CPU has been reloaded).
Onchip - Use this option to look for information if the card went through Cold Reset (Card reloaded entirely – Hard reset).
For example:
sysadmin-vm:F0_SC0# show controller ccc reset-history on
Possible completions:
onboard CCC Reset history in onboard EEPROM detail information
onchip On-chip reset history entries since last CCC Cold Reset
sysadmin-vm:F0_SC0# show controller ccc reset-history onchip location 0/0
****************************************************
*** On Chip Reset History for location 0/0 ***
****************************************************
TimeofDay : Tue Oct 20 17:17:40 2015
Uptime : 18 days 02:01:59 <--
Resets : 2
Reset Reset Reset
idx Source Command Time
--- --------------- ------------ -------------------
0 ColdRst AssrtHR 2015/10/02 15:15:43
1 ColdRst DeAssrtHR 2015/10/02 15:15:50 --> List reset source as "ColdRst"
sysadmin-vm:F0_SC0# show controller ccc reset-history onboard location 0/0
*****************************************************
*** On Board Reset History for location 0/0 ***
*****************************************************
Scratch EEPROM Magic : PON
Scratch EEPROM Version : 0x00014000
Reset History Magic : HIST
Number of Resets : 102 <---
Reset Reset Reset
idx Source Command Time
--- --------------- ------------ -------------------
0 ColdRst AssrtHR 1970/01/01 0:00:00 <-- ColdRst
1 ColdRst DeAssrtHR 1970/01/01 0:00:06
2 ColdRst AssrtHR 1970/01/01 0:00:00
3 ColdRst DeAssrtHR 1970/01/01 0:00:06
4 WarmRst AssrtHR 1970/05/03 7:21:55 <-- WarmRst
<output omitted>
ColdRst can either be initiated by Software or Hardware. WarmRst, is only initiated through Software. Other reset sources can be HRESET_L, SRESET_L, Wtchdog, SW_assgn or plain Rsrvd.
In addition, there are two unique entries per reset operation performed. An AssrtHR operation and a DeAssrtHR operation. This implies that a reset signal has been asserted and then de-asserted, therefore, the chip will complete the reset.
Take note of the timestamps of each of these operations. This CLI can be combined with the CCC Inventory Status CLI to determine when the reset occurred and how long the card has been up or down.
Next, check CCC stage that card went through at the time of its reload process. Listed here are different state examples:
Card that booted correctly from CCC point of view:
sysadmin-vm:F0_SC0# show controller ccc event-history brief location 0/0
CCC Card Event History for: 0/0
Card Event History as seen by Master (0/RP1)
Current State: CARD_READY
DATE TIME (UTC) STATE EVENT
----- ------------ ---------------------- ------------------------------
10/02 15:16:55.234 WAIT_BOOT_IMAGE ev_boot_ssd_image
10/02 15:16:54.233 BIOS_STARTED if_wait_ssd_image_booting
10/02 15:16:54.233 CPU_READY if_bios_started
10/02 15:16:54.231 OIR_INSERT_NOTIF if_cpu_is_ready
10/02 15:16:54.217 CCC_DRIVER_INIT if_oir_insert_notif_not_done
10/02 15:16:54.195 PON_POWERED_ON to_ccc_driver_init
10/02 15:16:54.195 CHECK_CCC_STATUS if_pon_powered_on
10/02 15:16:54.194 READ_IDPROM ev_idprom_available
10/02 15:16:53.942 GET_CCC_INFO ev_get_ccc_info_done
10/02 15:16:53.723 WAIT_ETH_READY ev_eth_available
10/02 15:16:52.560 CHECK_UBLAZE_BOOT ev_ublaze_boot_ok
10/02 15:16:52.539 WAIT_CCC_READY ev_ccc_ready
10/02 15:16:52.537 IDLE ev_presence_scan
Card currently in PXE_BOOTING stage:
sysadmin-vm:0_RP0# show controller ccc event-history brief location 0/3
CCC Card Event History for: 0/3
Current State: PXE_BOOTING
DATE TIME (UTC) STATE EVENT
----- ------------ ---------------------- ------------------------------
08/07 19:50:40.607 BIOS_STARTED if_internal_pxe_booting
08/07 19:50:40.607 WAIT_BIOS_START ev_bios_started
08/07 19:50:18.605 CPU_READY if_bios_not_started
08/07 19:50:18.595 CCC_DRIVER_INIT if_cpu_is_ready
08/07 19:50:18.568 PON_POWERED_ON to_ccc_driver_init
08/07 19:50:18.568 CHECK_CCC_STATUS if_pon_powered_on
08/07 19:50:18.567 GET_CCC_INFO ev_get_ccc_info_done
08/07 19:50:18.550 WAIT_ETH_READY ev_eth_ready
08/07 19:50:18.550 CHECK_UBLAZE_BOOT ev_ublaze_boot_ok
08/07 19:50:18.517 PON_UP_WARM ev_ccc_reset_done
08/07 19:50:12.627 PON_DOWN_WARM ev_pon_up_warm
08/07 19:50:08.239 PON_DOWN_WARM ev_warm_reset_req_ignored
08/07 19:50:07.239 PON_DOWN_WARM ev_warm_reset_req_ignored
08/07 19:50:06.239 PON_DOWN_WARM ev_warm_reset_req_ignored
08/07 19:50:05.239 PON_DOWN_WARM ev_warm_reset_req_ignored
08/07 19:50:04.238 PON_DOWN_WARM ev_warm_reset_req_ignored
Card unable to boot due to chip stuck in GET_CCC_INFO:
sysadmin-vm:0_RP0# show controller ccc event-history brief location 3/6
CCC Card Event History for: 3/6
Card Event History as seen by Master (3/RP0)
Current State: GET_CCC_INFO
DATE TIME (UTC) STATE EVENT
----- ------------ ---------------------- ------------------------------
10/26 23:43:04.559 UBLAZE_NOT_READY ev_timer_expired
10/26 23:42:34.559 CHECK_UBLAZE_BOOT ev_timer_expired
10/26 23:42:24.528 WAIT_CCC_READY ev_ccc_ready
10/26 23:42:21.516 RECOVERY_RESET ev_timer_expired
10/26 23:42:03.516 CHECK_UBLAZE_BOOT ev_ublaze_pre_boot_failed
10/26 23:41:52.480 WAIT_CCC_READY ev_ccc_ready
10/26 23:41:49.468 RECOVERY_RESET ev_timer_expired
10/26 23:41:32.467 WAIT_CCC_READY ev_no_fpga_ok_signal
10/26 23:41:29.456 RECOVERY_RESET ev_timer_expired
10/26 23:41:13.455 WAIT_CCC_READY ev_no_fpga_ok_signal
10/26 23:41:10.444 RECOVERY_RESET ev_timer_expired
10/26 23:40:55.444 CHECK_UBLAZE_BOOT ev_ublaze_pre_boot_failed
10/26 23:40:55.439 WAIT_CCC_READY ev_ccc_ready
10/26 23:40:52.320 IDLE ev_presence_scan
Card unable to boot because of POWER_UP_FAILED state:
sysadmin-vm:0_RP0# show controller ccc event-history brief location 0/2
CCC Card Event History for: 0/2
Current State: POWER_UP_FAILED
DATE TIME (UTC) STATE EVENT
----- ------------ ---------------------- ------------------------------
08/05 14:55:17.449 POWER_UP_FAILED ev_wdog_timeout
08/05 14:45:31.265 CCC_DRIVER_INIT if_pwr_up_failed
08/05 14:45:31.260 CHECK_CCC_STATUS if_pwr_up_failed_again
08/05 14:45:31.258 GET_CCC_INFO ev_get_ccc_info_done
08/05 14:45:31.223 WAIT_ETH_READY ev_eth_ready
08/05 14:45:31.157 CHECK_UBLAZE_BOOT ev_ublaze_boot_ok
08/05 14:45:31.124 PON_UP_WARM ev_ccc_reset_done
08/05 14:45:17.489 CCC_IN_RESET ev_pon_up_warm
08/05 14:45:08.921 POWER_UP_FAILED ev_pon_down_warm
08/05 14:35:07.152 POWER_UP_FAILED ev_wdog_timeout
08/05 14:25:20.946 CCC_DRIVER_INIT if_pwr_up_failed
08/05 14:25:20.941 CHECK_CCC_STATUS if_pwr_up_failed_again
08/05 14:25:20.939 GET_CCC_INFO ev_get_ccc_info_done
08/05 14:25:20.923 WAIT_ETH_READY ev_eth_ready
08/05 14:25:20.887 CHECK_UBLAZE_BOOT ev_ublaze_boot_ok
08/05 14:25:20.830 PON_UP_WARM ev_ccc_reset_done
Aborted: by user
If you use this command with the brief option, it does not give complete data related to the root cause of the issues. For that information replace the brief keyword with detail.
Note: This is the most important CLI when troubleshooting cards do not boot on the CCC Level.
Focus on the Event Desc and Reason of the failure to get a better explanation on the failure.
sysadmin-vm:F0_SC0# show controller ccc event-history detail location 0/0
CCC Card Event History for: 0/0
Card Event History as seen by Master (0/RP1)
Event buffer info:
Total number of events recorded: 13
Number of events available for display: 13
Current State: CARD_READY
EVENT #: 12 (record index = 12)
TIMESTAMP: 2015/10/02 15:16:55.234814 UTC
STATE: WAIT_BOOT_IMAGE
EVENT: ev_boot_ssd_image
EVENT DESC: SSD image is booting
EVENT #: 11 (record index = 11)
TIMESTAMP: 2015/10/02 15:16:54.233898 UTC
STATE: BIOS_STARTED
EVENT: if_wait_ssd_image_booting
EVENT #: 10 (record index = 10)
TIMESTAMP: 2015/10/02 15:16:54.233855 UTC
STATE: CPU_READY
EVENT: if_bios_started
EVENT #: 9 (record index = 9)
TIMESTAMP: 2015/10/02 15:16:54.231426 UTC
STATE: OIR_INSERT_NOTIF
EVENT: if_cpu_is_ready
EVENT #: 8 (record index = 8)
TIMESTAMP: 2015/10/02 15:16:54.217351 UTC
STATE: CCC_DRIVER_INIT
EVENT: if_oir_insert_notif_not_done
EVENT #: 7 (record index = 7)
TIMESTAMP: 2015/10/02 15:16:54.195808 UTC
STATE: PON_POWERED_ON
EVENT: to_ccc_driver_init
EVENT #: 6 (record index = 6)
TIMESTAMP: 2015/10/02 15:16:54.195786 UTC
STATE: CHECK_CCC_STATUS
Here are example outputs of different scenarios.
Card that didn't boot due to power problems and got stuck in POWER_UP_FAILED:
Pay attention to ERROR_INFO to get details about the failure.
sysadmin-vm:0_RP0# show controller ccc event-history detail location 0/2
CCC Card Event History for: 0/2
Event buffer info:
Total number of events recorded: 692
Number of events available for display: 255
Current State: POWER_UP_FAILED
EVENT #: 691 (record index = 179)
TIMESTAMP: 2014/08/05 14:55:17.449979 UTC
STATE: POWER_UP_FAILED
EVENT: ev_wdog_timeout
EVENT DESC: CCC watchdog timeout event
ERROR INFO: wdog__0 SysAdmin VM Watchdog stage1:0
<output omitted>
Card that didn't boot and got stuck in CCC_NOT_READY (chip problem):
sysadmin-vm:0_RP0# show controller ccc event-history detail location 0/FC2
CCC Card Event History for: 0/FC2
Event buffer info:
Total number of events recorded: 2
Number of events available for display: 2
Current State: CCC_NOT_READY
EVENT #: 1 (record index = 1)
TIMESTAMP: 2014/08/04 14:10:49.891845 UTC
STATE: WAIT_CCC_READY
EVENT: ev_ccc_ready_timeout
EVENT DESC: Timeout waiting for CCC to be ready
ERROR INFO: CCC READY Timeout - CLOCK_OK signal not being asserted (I/O Expander port0=0xf0, port1=0xff) ß
<output omitted>
There are instances where Cards needs to be removed/re-inserted. For that, CCC component provides OIR-History for given Rack tracking card insertion/removal. Note the keyword slave is used. This will give information on the standby RP.
sysadmin-vm:0_RP0# show controller ccc oir-history rack 0
Cards OIR History of rack: 0
OIR Events as seen by Master (0/RP0)- View from the Active RP
DATE TIME (UTC) EVENT LOC CARD TYPE SERIAL NO
----- ------------ ---------- ----- ------------------ -----------
10/09 16:59:14.280 INSERTED 0/0 NC6-10X100G-M-K SAL1650UCN9
10/09 16:58:49.064 REMOVED 0/0 NC6-10X100G-M-K SAL1650UCN9
<output omitted>
sysadmin-vm:0_RP1# show controller ccc slave oir-history rack 0
Cards OIR History of rack: 0
OIR Events as seen by Slave (0/RP1)- <-- View from the standby RP
DATE TIME (UTC) EVENT LOC CARD TYPE SERIAL NO
----- ------------ ---------- ----- ------------------ -----------
11/06 05:54:31.374 DISCOVERED 0/2 NC6-10X100G-M-K SAD161300XK
11/06 05:53:37.442 DISCOVERED 0/6 NC6-10X100G-M-K SAL1649TN46
<output omitted>
Information must be the same from both RP point of view.
Using this combination of commands helps to determine the root cause of why the Fabric Card was unable to boot.
Card Stuck in POWEROFF or PRESENT STATE
If the card is stuck in POWEROFF/PRESENT STATE, it is very likely that it went through multiple resets and was Powered-OFF by shelf_mgr.
Issue these commands in order to determine the root cause of the issue:
- Collect show tech ctrace from Sysadmin VM
- show platform detail location <>
- show reboot-history card location <> (find out how many times it went through the RESET process)
- show controller ccc event-history detail location <>
- show controller ccc reset-history onboard location <>
Card can be reset with the use of the hw-module reset command and booting process can be observed with the use of rconsole procedure explained later in this document.
sysadmin-vm:F0_SC0# hw-module location 0/0 reload
Reload hardware module ? [no,yes] Yes
Card Stuck in POWER_ON STATE
If the card is stuck POWERED_ON, it means that CCC has turned on the basic power zones that are needed for other Drivers to start their work. It is the responsibility of respective Drivers to move the card to an Operational state.
The SFE Driver moves FCs to Operational state, after it detects and initializes all its internal devices on that card also known as ASICs.
ESD Driver moves SC-SW cards and Scapa's LCs to Operational state (almost immediately, nothing to check/initialize unlike SFE Driver).
If Card is stuck in POWERED_ON state, it means that one of the above drivers had trouble to move the card to Operational state. Problem is seen more often on CPU-less cards. Ex: Fabric Cards or SC's Switch cards (SC-SW).
The first step is to check ccc event-history command:
sysadmin-vm:F0_SC0# show controller ccc event-history detail location 0/0
CCC Card Event History for: 0/0
Card Event History as seen by Master (0/RP1)
Event buffer info:
Total number of events recorded: 13
Number of events available for display: 13
<output ommited>
EVENT #: 7 (record index = 7)
TIMESTAMP: 2015/10/02 15:16:54.195808 UTC
STATE: PON_POWERED_ON
EVENT: to_ccc_driver_init
Next, validate the basic power zones needed:
sysadmin-vm:0_RP0# show controller ccc register location 0/RP0 offset 0x4c
Register Register
Address Value
------------------------
0x4C 0x3 - zones 0 and 1 OK
sysadmin-vm:0_RP0# show controller ccc register location 0/RP0 offset 0x50
Register Register
Address Value
------------------------
0x50 0x3 - zone 0 and 1 Enabled
sysadmin-vm:0_RP0# show controller ccc power detail location 0/RP0
Power detail : Zone information for 0/RP0:
---------------------------------------------------------
| Power Zone | Power Status | Power Contrl | Power Fault |
---------------------------------------------------------
| 0 | OK | SET | -- | - Power Status OK
| 1 | OK | SET | -- | - Power Status OK
sysadmin-vm:F0_SC0# show controller ccc i2c-dev ioexpander location 0/0
CCC IO Expander information for location: 0/0
------------------------------------------------------
Port 0: 0x3e
Port Bit I/O Val Bit Name
-------------------------------------
P0 0 O 0 Power Cycle
P0 1 I 1 FPGA OK
P0 2 I 1 uBlaze OK
P0 3 I 1 Clock OK
P0 4 I 1 Core Volt OK
P0 5 I 1 OTH Volt0 OK
P0 6 I 0 OTH Volt1 NOT OK
P0 7 I 0 OTH Volt2 NOT OK
Port 1: 0x3
Port Bit I/O Val Bit Name
-------------------------------------
P1 0 I 1 FPGA INIT OK
If this validation does not lead to a root cause then next step would be to open a TAC Service Request.
Card Stuck in SW_INACTIVE STATE
sysadmin-vm:0_RP0# show platform
Location Card Type HW State SW State Config State
--------------------------------------------------------------------- -------
0/1 P-L-10X100G-F-P POWERED_OFF SW_INACTIVE SHUT
0/RP0 P-L-RP OPERATIONAL OPERATIONAL NSHUT
0/RP1 P-L-RP OPERATIONAL OPERATIONAL NSHUT
0/FC0 P-L-FC-S OPERATIONAL N/A NSHUT
0/FC1 P-L-FC-S OPERATIONAL N/A NSHUT
0/FT0 PANINI-SIM-FT OPERATIONAL N/A NSHUT
0/FT1 PANINI-SIM-FT OPERATIONAL N/A NSHUT
Possible failure reasons are:
- Host OS not booting due to SSD access issue
- Host OS booting stuck due to HW issue
- SysAdmin VM failed to get spawned
- Control Ethernet connection issues
- Card MAC/IP not programmed due to SW failure
- Ethernet switches not getting programmed correctly by CCC Power-On interpreter
- ESD Color Switch image failed to get programmed in the CCC SPI flash
Scenario 1. SW_EVENT_FAILURE: SW_EVENT_ADMIN_VM_FAILURE Reported by shelf_mgr syslogs
sysadmin-vm:0_RP0# show platform
Location Card Type HW State SW State Config State
----------------------------------------------------------------------------
0/1 PROTO-CXP-2XPITA OPERATIONAL OPERATIONAL NSHUT
0/RP0 NC6-RP OPERATIONAL OPERATIONAL NSHUT
0/RP1 NC6-RP OPERATIONAL SW_INACTIVE NSHUT
0/FC0 NC6-FC-MC OPERATIONAL N/A NSHUT
0/CI0 P-L-CRFT OPERATIONAL N/A NSHUT
0/FT0 P-L-FANTRAY OPERATIONAL N/A NSHUT
0/FT1 P-L-FANTRAY OPERATIONAL N/A NSHUT
There could be several different reasons as to why RP1 does not boot. The easiest way to find out the issue is to rconsole on the RP and check the logs on (refer to rconsole procedure that is a the bottom of this document).
Scenario 2. LC Stuck in FAILED State, Last Event: HW_EVENT_FAILURE fail_code=LC_POWER_MAIN_FAULT
Ensure that the HW state shows FAILED and SW State shows SW_INACTIVE:
sysadmin-vm:0_RP0# show platform location 0/1
Location Card Type HW State SW State Config State
----------------------------------------------------------------------------
0/1 NC6-60X10GE-M-S FAILED SW_INACTIVE NSHUT
Execute this command and check Last Even Reason:
sysadmin-vm:0_RP0# show platform detail location 0/1
Platform Information for 0/1
PID : NC6-60X10GE-M-S
Description : "NCS 6000 60x10G Multi-Service SFP+"
VID/SN : V01
HW Oper State : FAILED
SW Oper State : SW_INACTIVE
Configuration : "NSHUT RST"
HW Version : 0.6
Last Event : HW_EVENT_FAILURE
Last Event Reason : "pon exit <-- UP_WARM_RESET cnt=123 fail_code=LC_POWER_MAIN_FAULT"
sysadmin-vm:0_RP0#
Filter syslog for the affected card in order to check log messages:
0/RP0/ADMIN0:Jun 21 00:33:13.487 : cm[1795]: %ROUTING-TOPO-5-OIR_ACTION : OIR card failed having serial number: SAD173501R7.
0/RP0/ADMIN0:Jun 21 00:33:13.528 : shelf_mgr[1818]: %INFRA-SHELF_MGR-5-CARD_INSERTION : Location: 0/1, Serial #: SAD173501R7
0/RP0/ADMIN0:Jun 21 00:33:13.528 : shelf_mgr[1818]: %INFRA-SHELF_MGR-6-HW_EVENT : Rcvd HW event HW_EVENT_FAILURE, event_reason_str 'Initial discovery FAIL: EXIT0, power request on , but not finish ccc-pon startup. power_control 0x00000001' for card 0/1
0/RP0/ADMIN0:Jun 21 00:33:13.530 : shelf_mgr[1818]: %INFRA-SHELF_MGR-3-CARD_HW_FAILED : Card: 0/1 hardware state going to FAILED
0/RP0/ADMIN0:Jun 21 00:34:06.734 : shelf_mgr[1818]: %INFRA-SHELF_MGR-6-HW_EVENT : Rcvd HW event HW_EVENT_RESET, event_reason_str 'pon enter --> DOWN_WARM_RESET cnt=3! ' for card 0/1
0/RP0/ADMIN0:Jun 21 00:34:15.987 : shelf_mgr[1818]: %INFRA-SHELF_MGR-6-HW_EVENT : Rcvd HW event HW_EVENT_POWERED_OFF, event_reason_str 'CCC Warm Reset #8' for card 0/1
0/RP0/ADMIN0:Jun 21 00:34:21.419 : cm[1795]: %ROUTING-TOPO-5-OIR_ACTION : OIR card failed having serial number: SAD173501R7.
0/RP0/ADMIN0:Jun 21 00:34:21.459 : shelf_mgr[1818]: %INFRA-SHELF_MGR-3-CARD_HW_FAILED : Card: 0/1 hardware state going to FAILED
0/RP0/ADMIN0:Jun 21 00:34:21.459 : shelf_mgr[1818]: %INFRA-SHELF_MGR-6-HW_EVENT : Rcvd HW event HW_EVENT_FAILURE, event_reason_str 'pon exit <-- UP_WARM_RESET cnt=4 fail_code=LC_POWER_MAIN_FAULT' for card 0/1
Execute show reboot history command and verify if there is a huge time difference between AssrtHR and DeAssrtHR messages. This is probably because there is an internal connectivity issue between the VM’s.
sysadmin-vm:0_RP0#show reboot-history card location 0/1
Reset history example: every 20 mins for 2 hours before it recovered:
33 0 WarmRst DeAssrtHR 0x00000F32 0x53A4D367 Sat Jun 21 00:35:51 2014
34 0 WarmRst AssrtHR 0x00000F10 0x53A4D81D Sat Jun 21 00:55:57 2014
35 0 WarmRst DeAssrtHR 0x00000F32 0x53A4D821 Sat Jun 21 00:56:01 2014
36 0 WarmRst AssrtHR 0x00000F10 0x53A4DCD7 Sat Jun 21 01:16:07 2014
Card Stuck in UNKNOWN STATE
When the Card reports Admin state as UNKNOWN, most likely CCC could not read the IDPROM from the board, therefore, the card will not be able to complete the boot. For such instances, execute these commands on the given location:
RP/0/RP0/CPU0:A41-PE1#show platform
Node name Node type Node state Admin state Config state
-----------------------------------------------------------------------------------
0/RP1 NC6-RP OPERATIONAL UNKNOWN
0/FC1 NC6-FC OPERATIONAL UNKNOWN
sysadmin-vm:F0_SC0# show controller ccc event-history brief location 0/0
CCC Card Event History for: 0/0
Card Event History as seen by Master (0/RP1)
Current State: CARD_READY
DATE TIME (UTC) STATE EVENT
----- ------------ ---------------------- ------------------------------
10/02 15:16:54.194 READ_IDPROM ev_idprom_available
10/02 15:16:53.942 GET_CCC_INFO ev_get_ccc_info_done
10/02 15:16:53.723 WAIT_ETH_READY ev_eth_available
10/02 15:16:52.560 CHECK_UBLAZE_BOOT ev_ublaze_boot_ok
10/02 15:16:52.539 WAIT_CCC_READY ev_ccc_ready
10/02 15:16:52.537 IDLE ev_presence_scan
Next step is to verify if CCC runs on the LC at the time of the bootup process with the use of rconsole:
1. Collect the process ID in SysadminVM:
sysadmin-vm:F0_SC0# show processes ccc_driver location 0/0
----------------------------------------------------------------------
PID: 2525
Executable path: /opt/cisco/calvados/packages/ncs6k-sysadmin-boot-5.2.4.CSCut24295 .all-1.0.0/sbin/ccc_driver
Instance #: 0
Respawn: ON
Respawn count: 1
Max. spawns per 4 mins: 4
Last started: 10/02/2015 15:17:23.000
Process state: Run
startup_path: /opt/cisco/calvados/packages/ncs6k-sysadmin-boot-5.2.4.CSCut24295 .all-1.0.0/etc/startup/ccc_driver.startup
Ready: 5s
2. Rconsole to the LC with the use of these commands:
sysadmin-vm:F0_SC0# attach location 0/RP0
[sysadmin-vm:0_RP0:~]$ exec chvrf 2 bash
[sysadmin-vm:0_RP0:~]$ chvrf 0 bash
[sysadmin-vm:0_RP0:~]$ /opt/cisco/calvados/sbin/rconsole -l 0/0
Connecting to location 0/0 (backplane-slotid 16, console 0)
Escape sequence is "end"
Waiting for card info from CCC-driver for slot 16
Got card info from CCC-driver for slot 16
IOS Build Date : 04/22/2015 by lchinnad
System Memory Speed : 1334 MHz
Processor Type : Intel(R) Xeon(R) CPU E5-2418L @ 2.00GHz
Press F12 to goto Boot Manager..
Booting System Host OS..
Waiting For CCC Valid Time of Day..
Waiting For CCC Valid Time of Day..
CCC Time: Fri Oct 2 15:16:54 2015
GNU GRUB version 2.00
Press F2 to goto grub Menu..
Booting from Disk..
Loading Kernel..
Loading initrd..
[ 1.949229] i8042: No controller found
Starting udev: [ OK ]
Switching to new root and running init.
Starting udev: [ OK ]
Actual changes:
large-receive-offload: off [requested on]
ntuple-filters: on
Setting hostname host: [ OK ]
Checking filesystems:[ OK ]
Entering non-interactive startup
Bringing up loopback interface: [ OK ]
Bringing up interface eth0: Device eth0 does not seem to be present, delaying initialization.
[FAILED]
Starting system logger: [ OK ]
Starting kernel logger: [ OK ]
Starting kdump:[ OK ]
Starting system message bus: [ OK ]
Starting smartd: [ OK ]
Generating SSH1 RSA host key: [ OK ]
Generating SSH2 RSA host key: [ OK ]
Generating SSH2 DSA host key: [ OK ]
Starting sshd: [ OK ]
Starting xinetd: [ OK ]
Starting crond: [ OK ]
Starting libvirtd daemon: [ OK ]
Starting NCS6k programs for LC on hostos: [ OK ]
mcelog start/running, process 2637
Creating default host password file
serial (/dev/ttyserial (/dev/ttyS1) start/running, process 2649
host login: root
Password:
[host:~]$
[host:~]$
[host:~]$ telnet 0 50001 <-- to get to Calvados
Trying 0.0.0.0...
Connected to 0.
Escape character is '^]'.
sysadmin-vm:0_0 login:
sysadmin-vm:0_0 login: root
Password:
[sysadmin-vm:0_0:~]$ pgrep ccc <- use pgrep to check if the process is running
2525
[sysadmin-vm:0_0:~]$ exit
logout
Use RCONSOLE Tool
When cards are unable to boot, The NCS6008 provides a built-in feature which gives the capability to perform remote console on the card and see the reason why the card is stuck and unable to boot. This feature is called RCONSOLE and here is an example of its usage.
Procedure to Rconsole on specific LC:
- Navigate to SysadminVM
- Attach to active RP.
- Change to global VRF chvrf 0 bash
- Execute /opt/cisco/calvados/sbin/rconsole -l (Linecard)
Example:
RP/1/RP1/CPU0:6008-B#admin
sysadmin-vm:F0_SC0#
sysadmin-vm:F0_SC0# attach location 0/RP0 <-- You must be connected to the RP's to be able to rconsole
Tue Oct 20 18:23:54.740 UTC
[sysadmin-vm:0_RP0:~]$ exec chvrf 2 bash
[sysadmin-vm:0_RP0:~]$ chvrf 0 bash
[sysadmin-vm:0_RP0:~]$ /opt/cisco/calvados/sbin/rconsole -l 0/0 ß This is LC 0/0
Connecting to location 0/0 (backplane-slotid 16, console 0)
Escape sequence is "end"
Waiting for card info from CCC-driver for slot 16
This procedure is widely used by TAC to determine the current state of the card and check where its stuck.
List of Commands to be Collected before you Open a TAC Case
XR VM:
Show platform
show install active
Show version
Show reboot history location 0/0/cpu0
dir misc/disk1
show cli history detail
Show log
show tech-support npu
SysAdmin VM:
show platform detail
show platform slices
show install active
show sdr default-sdr reboot-history
show reboot-history card location <>
show controller ccc reset-history onbo loc <>
show controller ccc reset-history onch loc <>
show controller ccc event-history detail location <>
show tech-support ccc
show tech-support Hbloss
show tech-support slice_manager
show tech-support ctrace
show tech-support sdr_mgr
show log
Contributed by Cisco Engineers
- Andre PedroCisco TAC Engineer
FeedbackContact Cisco
- Open a Support Case
- (Requires a Cisco Service Contract)