Symptom

- Tam services is not able to initialize - "type6_server" continuous restart RP/0/RP0/CPU0: tamsvcs_tamm[216]: %SECURITY-TAMSVCS-3-NOT_INITIALIZED : Tam services is not able to initialize RP/0/RP0/CPU0: processmgr[51]: %OS-SYSMGR-7-DEBUG : tamsvcs_tamm(1) (jid 216) did not signal end of initialization RP/0/RP0/CPU0: processmgr[51]: type6_server(1) (jid 393) (pid 11760) (fail_count 5) abnormally terminated, restart scheduled RP/0/RP0/CPU0: processmgr[51]: %OS-SYSMGR-3-ERROR : type6_server(1) (jid 393) exited, will be respawned with a delay (slow-restart) RP/0/RP0/CPU0: processmgr[51]: %OS-SYSMGR-3-ERROR : type6_server(393) (fail count 5) will be respawned in 120 seconds RP/0/RP0/CPU0: processmgr[51]: %OS-SYSMGR-7-DEBUG : type6_server(1) (jid 393) did not signal end of initialization "Failed to read master file error TAM_ERROR_CRYPTO_INTEGRITY_FAIL" can be seen from tam_debug.log Issue occur due to tam data corruption. Journaling support will avoid data corruption.

Conditions

powercycles, reload, driver failures, abrupt process kills could cause TAM-DB corruption.

Workaround

Four-step recovery if TAM errors are seen on RP-Card: 1) Shut down tamsvcs_tamm process on RP, with the command: "process shutdown tamsvcs_tamm location " 2) Delete TAM-DB from that particular node, by executing command: " run rm -rf /misc/config/SWTAmRoot " 3) Start tamsvcs_tamm process on RP, with a command: "process start tamsvcs_tamm location " 4) Since data is lost on main-RP, one needs to Re-add to key-chain, any passwords (like MACSec preshared keys) which previously might have been to key-chain earlier. This will help applications (like MACSec) to fetch back their pre-shared keys./passwords and function as normal. Three-step recovery if issue seen on Line-Card: 1) Shut down tamsvcs_tamm process on particular node, with the command: "proc shut tamsvcs_tamm location " 2) Delete TAM-DB from that particular node, by attaching to console of that node/LC and running rm command: "attach location " "rm -rf /misc/config/SWTAmRoot " 3) Start tamsvcs_tamm process on that particular node, with command: "proc start tamsvcs_tamm location "

Further Problem Description