#620 Report cpu blade Mezznine card, and memory details
OA SOAP plug-in needs enhancements to report Mezzanine cards and memory on CPU
Blades. Mezzanine card information includes port numbers and WW IDs for fiber
channel Mezz cards. Using FC card as an example, for Ethernet relevant details
would include speed etc. This functionality is important for Kevlar and also as
a complete support for c-class servers.
How do you obtain mezzanine card and memory details: (available now)
The SOAP calls are getBladePortMap and getBladePortMapArray.
You can only get the port info for present blades however, here is a snippet
for an Ethernet device:
<hpoa:port>
<hpoa:portNumber>2</hpoa:portNumber>
<hpoa:wwpn>00:16:35:C6:45:0C</hpoa:wwpn>
<hpoa:fabric>FABRIC_TYPE_ETH</hpoa:fabric>
<hpoa:status>FABRIC_STATUS_OK</hpoa:status>
<hpoa:extraData
hpoa:name="iSCSIwwpn">00:16:35:C6:45:0D</hpoa:extraData>
</hpoa:port>
This info can be seen on the GUI port mapping pages as well as on the blade
info page.
How do you obtian cpu memory details:
getBladeInfo will give you the blade CPU info, and memory total.
Below is the some information from PG's notes. This may be helpful in the future.
Blade
CPU ZONE
---------
If the CPU is dual (or more) core, then we can have CPU ZONE and individual
CPUs will be part of the CPU zone.
getBladeThermalInfoArray SOAP call
bladeThermalInfo->entityId = 3 (CPU/CPU Zone)
If bladeThermalInfo->description has "CPU ZONE"
the sensor information for CPU zone is available.
bladeThermalInfo->criticalThreshold
bladeThermalInfo->cautionThreshold
bladeThermalInfo->temperatureC
NOTE: store the entityInstance in rpt for future use.
CPU
-----
getBladeInfo SOAP call
bladeInfo->numberOfCpus
bladeInfo->cpus->bladeCpuInfo[n]->cpuType
badeInfo->cpus->bladeCpuInfo[n]->cpuSpeed
getBladeThermalInfoArray SOAP call gives the more details on the
presence/absence of CPUs and also type of CPUs (dual/quad etc).
It also gives the sensor information for each CPUs
getBladeThermalInfoArray SOAP call
bladeThermalInfo->entityId = 9 (CPU ZONE)
If bladeThermalInfo->description has "CPU 1/2/3"
the sensor information for CPU 1 is available.
bladeThermalInfo->criticalThreshold
bladeThermalInfo->cautionThreshold
bladeThermalInfo->temperatureC
NOTE: store the entityInstance in rpt for future use.
Group the responses of CPU 1/2 under 'CPU ZONE'.
The entity path of the CPU 1/2 will be under the CPU ZONE.
NIC
-----
bladeInfo->numberOfNics
bladeInfo->nics->bladeNicInfo[n]->port
bladeInfo->nics->bladeNicInfo[n]->macAddress
DIMM
-------
getBladeDimmInfoArray SOAP call
bladeDimmInfoArray->bladeDimmInfo[n]->speedMHz
bladeDimmInfoArray->bladeDimmInfo[n]->sizeMb
bladeDimmInfoArray->bladeDimmInfo[n]->slot
Memory
-----
bladeInfo->memory (gives the size of the RAM).
getBladeThermalInfoArray (Memory zone) gives the temperature sensor information
of memory zone.
bladeThermalInfo->entityId = 8 (CPU)
If bladeThermalInfo->description has "CPU 1/2/3"
the sensor information for CPU 1 is available.
bladeThermalInfo->criticalThreshold
bladeThermalInfo->cautionThreshold
bladeThermalInfo->temperatureC
Mezzanine cards
----------------
getBladeMezzInfoEx SOAP call (supported from 2.25)
bladeMezzInfoEx->mezzInfo[n]->mezznumber
bladeMezzInfoEx->mezzInfo[n]->port[m]->portNumber
bladeMezzInfoEx->mezzInfo[n]->port[m]->guid[l]->type
bladeMezzInfoEx->mezzInfo[n]->port[m]->guid[l]->physicalFunction
getBladePortMap SOAP call
bladePortMap->status ==> status of the blade's port
bladePortMap->bladeSizeType ==> size of the blade
bladePortMap->numberOfMezzes
bladePortMap->mezz[n]->mezzNumber
bladePortMap->mezz[n]->mezzSlots->type
bladePortMap->mezz[n]->mezzSlots->slot
bladePortMap->mezz[n]->mezzDevices->name
bladePortMap->mezz[n]->mezzDevices->type
bladePortMap->mezz[n]->mezzDevices->status (currently coming as UNKNOWN)
bladePortMap->mezz[n]->mezzDevices->port[m]->portNumber
bladePortMap->mezz[n]->mezzDevices->port[m]->wwpn
bladePortMap->mezz[n]->mezzDevices->port[m]->fabric
bladePortMap->mezz[n]->mezzDevices->port[m]->status (currently coming as
UNKNOWN)
MP (iLO2)
----------
getBladeMpInfo SOAP call
bladeMpInfo->ipAdress
bladeMpInfo->macAdress
bladeMpInfo->dnsName
bladeMpInfo->modelName (can used for IDR)
bladeMpInfo->fwVersion (can used for IDR)
Or should we use the bladeStatus->diagnosticChecks->managementProcessorError
field for operational sensor.
<hpoa:managementProcessorError>NO_ERROR</hpoa:managementProcessorError>
Storage Blades
Memory (Disk zone)
We can support temperature sensor for the same.
Interconnect
NIC
---
getInterconnectTrayPortMap SOAP call
interconnectTrayPortMap->status ==> status of the interconnect
interconnectTrayPortMap->bladeSizeType ==> size of the interconnect
interconnectTrayPortMap->numberOfSlots ==> Is it number of slots occupied by
the interconnect
interconnectTrayPortMap->slot[n]->interconnectTraySlotNumber
interconnectTrayPortMap->slot[n]->portEnabled ..... (all other information may
not be valid)
CPU
----
getInterconnectTrayStatus SOAP call gives whether CPU fault has happened or
not.
We can configure CPU with RESOURCE and (operational) SENSOR capabilities set
OA
NIC
----
configureOaNicAuto => may configures to auto mode
configureOaNicForced => speed and duplex settings can be changed
configureOaStatic => ipAddress, netmask can be changed
Check what sensor and/or inventory information is available for every one of the above resources.
How many of the above things we need to model as HPI resource?
Feature added in checkin #7414