求助 一臺(tái)機(jī)器自動(dòng)宕機(jī) hacmp 產(chǎn)生force 事件 求原因

feitan

一臺(tái)機(jī)器優(yōu)于溫度過高, 宕機(jī)了, HACMP 產(chǎn)生 forced  events,        沒有切換   

想問下, HACMP 什么情況下會(huì)產(chǎn)生force 事件 而不產(chǎn)生 takeover 事件??

怎么設(shè)置才能當(dāng)一臺(tái) 宕機(jī)后, HACMP自動(dòng)產(chǎn)生takeover 自動(dòng)切換到備機(jī)上??  手動(dòng) takeover 切換沒問題的!!


謝謝!!!

HACMP Event Summary
Event: /usr/es/sbin/cluster/events/check_for_site_down p550a forced
Start time: Sun Sep  7 09:43:52 2008

End time: Sun Sep  7 09:43:52 2008

Action:                Resource:                        Script Name:
----------------------------------------------------------------------------
No resources changed as a result of this event
----------------------------------------------------------------------------

Sep  7 09:43:52 EVENT START: node_down p550a forced

:node_down[79] [[ high = high ]]
:node_down[79] version=1.45.1.1
:node_down[80] :node_down[80] cl_get_path
HA_DIR=es
:node_down[82] export NODENAME=p550a
:node_down[83] export PARAM=forced
:node_down[85] UPDATESTATDFILE=/usr/es/sbin/cluster/etc/updatestatd
:node_down[94] STATUS=0
:node_down[96] [[ -z  ]]
:node_down[97] EMULATE=REAL
:node_down[100] set -u
:node_down[102] ((  2 < 1  ))
:node_down[107] rm -f /tmp/.RPCLOCKDSTOPPED
:node_down[108] rm -f /usr/es/sbin/cluster/etc/updatestatd
:node_down[110] [[ forced = forced ]]
:node_down[110] [[ p550a = p550a ]]
:node_down[110] [[ REAL = EMUL ]]
:node_down[119] clchdaemons -r -d clstrmgr_scripts -t resource_locator
:node_down[124] cl_echo 3048 node_down : p550a forced down. Exiting immediately.\n node_down p550a
:cl_echo[49] version=1.13
:cl_echo[98] HACMP_OUT_FILE=/tmp/hacmp.out
Sep  7 2008 09:43:52 node_down : p550a forced down. Exiting immediately.
:node_down[125] return 0
Sep  7 09:43:53 EVENT COMPLETED: node_down p550a forced 0

yddll

hacmp在機(jī)器關(guān)機(jī)的時(shí)候執(zhí)行不執(zhí)行接管動(dòng)作，在/usr/es/sbin/cluster/etc/rc.shutdown里是會(huì)有定義的，可能你的環(huán)境問題導(dǎo)致的宕機(jī)應(yīng)驗(yàn)其中了的一種情況。

附/usr/es/sbin/cluster/etc/rc.shutdown部分內(nèi)容：

1. ###############################################################################
   
2. #   We're doing a shutdown because of an EPOW condition
   
3. ###############################################################################
   
4. if [[ $EPOW_SHUTDOWN == true ]] ; then
   
5.     ###########################################################################
   
6.     #   We attempt to do a 'graceful with takeover' shutdown, so that the
   
7.     #   resource groups get taken over when this node fails.  Note that this
   
8.     #   is in no sense guaranteed to complete before the node dies, but at
   
9.     #   least the other cluster nodes should perform takeover.
   
10.     ###########################################################################
    
11.     clstop -gr -y -s -N -S
    
12. ###############################################################################
    
13. #   A "fast shutdown" ("-F") was specified.  We turn this into a forced down
    
14. #   operation, mostly to preserve compatabilty
    
15. ###############################################################################
    
16. elif [[ $fast == true ]] ; then
    
17.     clstop -f -y -s -N -S
    
18. ###############################################################################
    
19. #   A "reboot" ("-r") was specified.  Do a graceful down, without takeover,
    
20. #   leaving the application cleanly down and HACMP out of the way on reboot.
    
21. ###############################################################################
    
22. elif [[ $reboot == true ]] ; then
    
23.     clstop -g -y -s -N -S
    
24. ###############################################################################
    
25. #   Every other shutdown is mapped to a graceful with takeover
    
26. ###############################################################################
    
27. else
    
28.     clstop -gr -y -s -N -S
    
29. fi
    
30. ###############################################################################
    
31. #   this is shutdown so remove the cluster shared memory
    
32. ###############################################################################
    
33. sleep 4
    
34. rm -f /usr/es/sbin/cluster/etc/ClSm 2>/dev/null

復(fù)制代碼

tyrone.dev

這種情況應(yīng)當(dāng)能切換，要不然HA就變成沒用的東西，查一下A機(jī)與B機(jī)的HA log,以及A機(jī)的B機(jī)的errpt

dongli

汗，溫度有多高��？都宕機(jī)了
查查日志吧