We are using Dell Open Manage SNMP System Health monitor on our servers. Many of our servers also have iDrac. What are the merits of using iDrac vs OpenManage ? There seems to be some differences in report results. For instance the OpenManage shows Last Down as 5 days, where iDrac shows 106 days for the same server. If there are any benefits to iDrac, is there a way to have it provide more accurate information ?
Hi and thank you for your KB-Post,
Regarding the following Dell/Poweredge-specific sensors in PRTG:
There's no easy answer, it will really depend on what you want to achieve and what alternative is available. Essentially, these are the main differences between the two approaches:
OMSA (Operating System)
This is the most common type of monitoring, it will usually be available but your operating System needs to support SNMP and integrate with Dell's OMSA to get vendor-specific sensors working. This approach allows you to not only monitor hardware-specific metrics but also deploy all other sorts of compatible sensors, for Disk Space, Memory utilization, CPU Usage, services and so on.
iDrac (Management Interface)
For devices that have a management interface, you can use this interface to monitor the server as well. This may have an easier set-up (since you configure the iDRAC and it is totally separated from the Server's OS) and doesn't require the operating system to have SNMP enabled/configured. As a matter of fact, you don't even need access to the operating server, which means that you can monitor an out-of-band server if the iDRAC is accessible. The downside of this approach is that for instance if the server is rebooted, turned off or disconnected from network, any sensors deployed on the iDRAC will keep working. This can be counter-intuitive and misleading, because the server will be unavailable for it's users but your monitor will still show everything as up.
It is definitively important to understand the ramifications of using the Out-off-band interface for monitoring. While you can monitor your servers via it's iDRAC (or similar management interface), I would suggest that you don't unless there's no other option. It doesn't by any means replace the regular monitoring using the server's address, but can be deployed as an "additional" safety measure and also to make sure that the management interface is available when needed (but a simple HTTP or Ping sensor on the iDRAC interface/address should suffice).
It makes a lot of sense to monitor a server via it's iDRAC in the following scenarios
- You're hosting the server but you don't have access to it's operating system, or it's in a different VLAN from which you can't poll it
- You don't want/can't install the Dell OMSA software on the server
- The server's operating system Doesn't support SNMP or doesn't integrate with OMSA.
Since the iDRAC and the Server have different addresses, you will have to deploy the server twice in PRTG, it could look like this
- Device: mail.corp.domain (actual server)
- SNMP System Uptime
- SNMP CPU Load
- Memory: Physical Memory
- Disk Free: C:\ Label: Serial Number 123456
- Disk Free: D:\ Label: Serial Number 123457
- Traffic - (011) Ethernet Team
- SMTP Sensor
- IMAP Sensor
- POP3 Sensor
- SNMP Dell PowerEdge Physical Disk Sensor 1
- SNMP Dell PowerEdge Physical Disk Sensor 2
- SNMP Dell PowerEdge Physical Disk Sensor 4
- SNMP Dell PowerEdge Physical Disk Sensor 5
- SNMP Dell PowerEdge System Health Sensor
- Ddevice: mail.corp.mgmt (iDRAC of mail.corp.domain)
Luciano Lingnau [Paessler Support]
PRTG failed to detect the CPU load and Memory uing SNMP in iDRAC. Only physical disks were detected. Is there a way to get the CPU load and memory usage using iDrac?
thank you for your inquiry.
I don't think you will be able to get CPU or Memory metrics via the iDRAC. The reason being that the "datasource" for these sensors is the HOST-RESOURCES-MIB, which might not be implemented on the iDRAC (but is implemented on Windows OS for example).
Luciano Lingnau [Paessler Support]