I'm checking my syslog settigns are all consistent and correct. Of course we can check and change them from vCenter (or new vCenter web client--shudder!), but more than a few hosts are best done by scripts/cli.
I can check this from PowerVCLI:
get-vmhost | Get-VMHostAdvancedConfiguration -Name Syslog.global.logHost
or specific hosts:
Get-VMHostAdvancedConfiguration -Name Syslog.global.logHost -VMHost host1
Name Value
---- -----
Syslog.global.logHost udp://1.2.3.4:514
and I can even fix it with a script using these commands
Set-VMHostAdvancedConfiguration -Name Syslog.global.logHost -Value 'udp://1.1.1.1:514' -VMHost host1
$esxcli.system.syslog.reload()
thanks to this cool script kindly provided by Caleb here.
So, the reloading of the syslog service. This isn't the usual linux service there's a command for reloading:
from the ESXi host (ssh/putty/DUI session):
~ # esxcli system syslog reload
the VMware kbase article(s) say to check the syslog is running with the good ol' linux ps command:
~ # ps | grep vmsyslogd
8666 8666 vmsyslogd /bin/python
8667 8666 vmsyslogd /bin/python
8668 8666 vmsyslogd /bin/python
But how do you verify whether either of these methods have worked (other than the absence of error messages?).
Seems /var/log/hostd.log is where this is recorded:
reload issued via PowerCLI:
2014-02-19T11:39:14.045Z [32B81B90 info 'Solo.VmwareCLI'] Dispatch reload
2014-02-19T11:39:14.114Z [32B81B90 info 'Solo.VmwareCLI'] Dispatch reload done
reload issued directly from ESXi linux login session:
2014-02-19T11:52:30.677Z [2A9DEB90 verbose 'Hostsvc.SyslogConfigProvider'] Running '/sbin/localcli system syslog config logger list'
2014-02-19T11:52:30.678Z [2A9DEB90 info 'SysCommandPosix'] ForkExec(/sbin/localcli) 393025
2014-02-19T11:52:30.830Z [2A9DEB90 verbose 'Hostsvc'] Received advanced config change notification
Can you shed any light on this?
Join the discussion and add your perspective. Thanks, KC
No comments:
Post a Comment