Hello community!
I'm trying to disable vmkernel.log debug logging but to no avail so far. My vmkernel.log is flooded by something like this:
[ACPI Debug] Integer 2015-06-30T07:17:08.051Z cpu0:32866)0x | 00000000 | 00001e00 |
[ACPI Debug] String 2015-06-30T07:17:08.051Z cpu0:32866)[0x16] "H2: 0xD7 Command == 0 "
[ACPI Debug] String 2015-06-30T07:17:08.051Z cpu0:32866)[0x0f] "H2: 0xD7 PETE: "
[ACPI Debug] Integer 2015-06-30T07:17:08.051Z cpu0:32866)0x | 00000000 | 00000080 |
[ACPI Debug] String 2015-06-30T07:17:08.051Z cpu0:32866)[0x1c] "H2: 0xD8 Notifying Processor"
[ACPI Debug] String 2015-06-30T07:17:08.052Z cpu0:32866)[0x24] "H2: 0xD3: Bit4 CSR_HA it is set to 0"
[ACPI Debug] String 2015-06-30T07:17:08.052Z cpu0:32866)[0x19] "H2: 0xD4 Bit3 CSR_HA -> 0"
[ACPI Debug] String 2015-06-30T07:17:08.055Z cpu0:32866)[0x28] "H2: 0xD0 Entering SCI Handler for HECI-2"
[ACPI Debug] String 2015-06-30T07:17:08.055Z cpu0:32866)[0x2d] "H2: 0xD1 Host Interrupt Status it is set to 1"
[ACPI Debug] String 2015-06-30T07:17:08.055Z cpu0:32866)[0x24] "H2: 0xD3: Bit4 CSR_HA it is set to 0"
[ACPI Debug] String 2015-06-30T07:17:08.055Z cpu0:32866)[0x19] "H2: 0xD4 Bit3 CSR_HA -> 0"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x28] "H2: 0xD0 Entering SCI Handler for HECI-2"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x2d] "H2: 0xD1 Host Interrupt Status it is set to 1"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x24] "H2: 0xD3: Bit4 CSR_HA it is set to 0"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x19] "H2: 0xD4 Bit3 CSR_HA -> 0"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x17] "H2: 0xD6 ME WP != ME RP"
[ACPI Debug] String 2015-06-30T07:17:08.350Z cpu0:32866)[0x19] "H2: 0xD6 Message header: "
[ACPI Debug] Integer 2015-06-30T07:17:08.351Z cpu0:32866)0x | 0000000080040011 |
[ACPI Debug] String 2015-06-30T07:17:08.351Z cpu0:32866)[0x26] "H2: 0xD6 TState:PState:SeqNo:Command: "
[ACPI Debug] Integer 2015-06-30T07:17:08.351Z cpu0:32866)0x | 00000000 | 00001f00 |
[ACPI Debug] String 2015-06-30T07:17:08.351Z cpu0:32866)[0x16] "H2: 0xD7 Command == 0 "
[ACPI Debug] String 2015-06-30T07:17:08.351Z cpu0:32866)[0x0f] "H2: 0xD7 PETE: "
[ACPI Debug] Integer 2015-06-30T07:17:08.351Z cpu0:32866)0x | 00000000 | 00000080 |
[ACPI Debug] String 2015-06-30T07:17:08.351Z cpu0:32866)[0x1c] "H2: 0xD8 Notifying Processor"
[ACPI Debug] String 2015-06-30T07:17:08.352Z cpu0:32866)[0x24] "H2: 0xD3: Bit4 CSR_HA it is set to 0"
[ACPI Debug] String 2015-06-30T07:17:08.352Z cpu0:32866)[0x19] "H2: 0xD4 Bit3 CSR_HA -> 0"
[ACPI Debug] String 2015-06-30T07:17:08.353Z cpu0:32866)[0x28] "H2: 0xD0 Entering SCI Handler for HECI-2"
[ACPI Debug] String 2015-06-30T07:17:08.353Z cpu0:32866)[0x2d] "H2: 0xD1 Host Interrupt Status it is set to 1"
[ACPI Debug] String 2015-06-30T07:17:08.353Z cpu0:32866)[0x24] "H2: 0xD3: Bit4 CSR_HA it is set to 0"
[ACPI Debug] String 2015-06-30T07:17:08.353Z cpu0:32866)[0x19] "H2: 0xD4 Bit3 CSR_HA -> 0"
What I did:
1. modified config.xml in /etc/vmware/hoistd. Currently <log> section looks like that:
<log>
<directory>/var/log/vmware/</directory>
<level>info</level>
<maxFileNum>8</maxFileNum>
<maxFileSize>524288</maxFileSize>
<name>hostd</name>
<outputToConsole>false</outputToConsole>
<outputToFiles>false</outputToFiles>
<outputToSyslog>true</outputToSyslog>
<syslog>
<facility>local4</facility>
<ident>Hostd</ident>
<logHeaderFile>/var/run/vmware/hostdLogHeader.txt</logHeaderFile>
</syslog>
</log>
I changed <level> to info (was verbose)
2. Changed advanced settings :
Config.HostAgent.log.level from verbose to info
Vpx.Vpxa.config.log.level from verbose to info.
No results even after host reboot.
Did I miss something?
This ESXi is Lenovo customized image.
Hi,
I met exactly same problem with our brand new Lenovo x3650 M5 machines. Unfortunately VMware was not able to say where those logs are coming from as there was no any hardware issue.
If you are feeling comfortable, you may decrease vmkernel loging level by:
esxcli system settings kernel set --setting=acpiDbgLevel --value=1
Then you need to reboot your host.
I hope it helps.
Hello,
Please check this KB article: VMware KB: Increasing VMware vCenter Server and VMware ESX/ESXi logging levels
Regards
Daniel
Hi, thanks for reply.
I already read this KB before posting. I have everything set to "info":
D:\> Get-VMHost -Name e133*| Get-AdvancedSetting -Name * |where value -Like info |ft -AutoSize
Name Value Type Description
---- ----- ---- -----------
Vpx.Vpxa.config.log.memoryLevel info VMHost
Config.HostAgent.log.level info VMHost
Vpx.Vpxa.config.log.level info VMHost
Vpx.Vpxa.config.level[SoapAdapter.HTTPService.HttpConnection].logLevel info VMHost
Vpx.Vpxa.config.level[SoapAdapter.HTTPService].logLevel info VMHost
"Verbose" not returning anything.
There is only one place where "verbose" is set in /etc/vmware/hostd/config.xml:
/etc/vmware/hostd # cat config.xml |grep -C 3 verbose
<path>libsolo.so</path>
<requires>hostsvc</requires>
<soapPort>8307</soapPort>
<traceAt>verbose</traceAt>
<webServer>
<port>8309</port>
</webServer>
Rest is set as "info"
No "verbose"in vpxa.cfg...
Hy please look to this article: vSphere Documentation Center
and this KB: VMware KB: Enabling trivia logging in VMware vCenter Server
You must have configured in the VC logging the Information option (Normal Logging)
Look in my config.xml in one VMware ESXi 5.5.0 build-1746018:
Hi, I verified vCenter settings - log level is "information"
MNy config.xml is slightly different from yours, but settings which you pointed are set like yours:
<level id="PropertyProvider">
<logLevel>info</logLevel>
<logName>PropertyProvider</logName>
<prefix>false</prefix>
</level>
<level id="Proxysvc">
<logLevel/>
<logName>Proxysvc</logName>
<prefix>true</prefix>
</level>
<level id="Snmpsvc">
<logLevel/>
<logName>Snmpsvc</logName>
<prefix>true</prefix>
</level>
<level id="SoapAdapter">
<logLevel>info</logLevel>
<logName>SoapAdapter.HTTPService</logName>
<prefix>true</prefix>
</level>
<level id="Statssvc">
<logLevel/>
<logName>Statssvc</logName>
<prefix>true</prefix>
</level>
<level id="Vcsvc">
<logLevel/>
<logName>Vcsvc</logName>
<prefix>true</prefix>
</level>
<level id="Vimsvc">
<logLevel/>
<logName>Vimsvc</logName>
<prefix>true</prefix>
</level>
<level id="Vmsvc">
<logLevel/>
<logName>Vmsvc</logName>
<prefix>true</prefix>
</level>
<log>
<directory>/var/log/vmware/</directory>
<level>info</level>
<maxFileNum>8</maxFileNum>
<maxFileSize>524288</maxFileSize>
<name>hostd</name>
<outputToConsole>false</outputToConsole>
<outputToFiles>false</outputToFiles>
<outputToSyslog>true</outputToSyslog>
<syslog>
<facility>local4</facility>
<ident>Hostd</ident>
<logHeaderFile>/var/run/vmware/hostdLogHeader.txt</logHeaderFile>
</syslog>
</log>
What configuration do yo have in the VC Logging ?? What version it is your VC?
Hi,
I met exactly same problem with our brand new Lenovo x3650 M5 machines. Unfortunately VMware was not able to say where those logs are coming from as there was no any hardware issue.
If you are feeling comfortable, you may decrease vmkernel loging level by:
esxcli system settings kernel set --setting=acpiDbgLevel --value=1
Then you need to reboot your host.
I hope it helps.
That done the trick, thanks!
Because I had to change this setting on bunch of hosts, I wrapped it in small PowerCLI script to run it on all hosts in specified cluster:
#sets ACPI debug in all hosts in $cluster host cluster
$cluster = "your_cluster_name_here"
$hosts = Get-Cluster -Name $cluster|Get-VMHost
$hosts |ForEach-Object {
$mycli = Get-EsxCli -VMHost $_.name
$mycli.system.settings.kernel.set("acpiDbgLevel","1")
}
Dzięki Janek :smileygrin:
The ACPI Debug messages were fixed in new Lenovo UEFI firmware:
https://www-947.ibm.com/support/entry/portal/docdisplay?lndocid=migr-5098203