When i enable network metric,i have received this error log

love6875520 commented 4 weeks ago

ERROR Error collecting metrics for host xx.xx.xxx.xxx: 3 error(s) occurred:

collected metric "idrac_network_port_health" { label:{name:"id" value:"NIC.Slot.1"} label:{name:"interface" value:"NIC.Slot.1"} label:{name:"status" value:""} gauge:{value:10}} was collected before with the same name and label values

I use idrac_exporter to collect more than 400 dell servers,when i enable network metric, more than 100 servers unable to collecte.

mrlhansen commented 4 weeks ago

Hi @love6875520

I need some additional information to start looking into this.

What machine models are causing problems? (e.g. Dell PowerEdge R740xd)
Can you run the exporter with the -verbose flag and send me the full output log when scraping one of the machines not working?

Based on this, I will probably need you to send me some JSON dumps from iDRAC later on.

love6875520 commented 4 weeks ago

The machine models include Dell PowerEdge R640 R650 R840

This is the verbose log,thx

collected metric "idrac_network_port_health" { label:{name:"id" value:"NIC.Slot.2"} label:{name:"interface" value:"NIC.Slot.2"} label:{name:"status" value:""} gauge:{value:10}} was collected before with the same name and label values
collected metric "idrac_network_port_speed_mbps" { label:{name:"id" value:"NIC.Slot.2"} label:{name:"interface" value:"NIC.Slot.2"} gauge:{value:0}} was collected before with the same name and label values
collected metric "idrac_network_port_link_up" { label:{name:"id" value:"NIC.Slot.2"} label:{name:"interface" value:"NIC.Slot.2"} label:{name:"status" value:""} gauge:{value:0}} was collected before with the same name and label values

mrlhansen commented 4 weeks ago

Thanks!

I have some R640 myself where the issue is not present, but it might depend the hardware and the version of iDRAC. I need two JSON dumps, which can be obtained using curl.

curl -k -u USERNAME https://xx.xx.xxx.xx/redfish/v1/Systems/System.Embedded.1/NetworkInterfaces/NIC.Slot.2
curl -k -u USERNAME https://xx.xx.xxx.xx/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkPorts/NIC.Slot.2-1

Replace the USERNAME with the username used for logging into iDRAC, and put the IP addresses back. The output should not contain anything sensitive, just state for the cards.

love6875520 commented 4 weeks ago

The machine info is below:

Server model：PowerEdge R640 BIOS Version：2.3.10 iDRAC firmware version 6.10.00.00

{"@odata.context":"/redfish/v1/$metadata#NetworkInterface.NetworkInterface","@odata.id":"/redfish/v1/Systems/System.Embedded.1/NetworkInterfaces/NIC.Slot.2","@odata.type":"#NetworkInterface.v1_2_1.NetworkInterface","Description":"Network Device View","Id":"NIC.Slot.2","Links":{"NetworkAdapter":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2"}},"Name":"Network Device View","NetworkDeviceFunctions":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkDeviceFunctions"},"NetworkPorts":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkPorts"},"Status":{"Health":null,"HealthRollup":null,"State":"Enabled"}}

{"@odata.context":"/redfish/v1/$metadata#NetworkPort.NetworkPort","@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkPorts/NIC.Slot.2-1","@odata.type":"#NetworkPort.v1_4_1.NetworkPort","ActiveLinkTechnology":"Ethernet","AssociatedNetworkAddresses":["14:02:EC:8D:D2:D4","14:02:EC:8D:D2:D5"],"CurrentLinkSpeedMbps":0,"Description":"Network Port View","EEEEnabled":null,"FlowControlConfiguration":null,"FlowControlStatus":null,"Id":"NIC.Slot.2","LinkStatus":null,"Name":"Network Port View","NetDevFuncMaxBWAlloc":[{"MaxBWAllocPercent":null,"NetworkDeviceFunction":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkDeviceFunctions/NIC.Slot.2-1"}},{"MaxBWAllocPercent":null,"NetworkDeviceFunction":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkDeviceFunctions/NIC.Slot.2-2"}}],"NetDevFuncMaxBWAlloc@odata.count":2,"NetDevFuncMinBWAlloc":[{"MinBWAllocPercent":null,"NetworkDeviceFunction":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkDeviceFunctions/NIC.Slot.2-1"}},{"MinBWAllocPercent":null,"NetworkDeviceFunction":{"@odata.id":"/redfish/v1/Chassis/System.Embedded.1/NetworkAdapters/NIC.Slot.2/NetworkDeviceFunctions/NIC.Slot.2-2"}}],"NetDevFuncMinBWAlloc@odata.count":2,"Oem":{},"PhysicalPortNumber":"2","Status":{"State":"Enabled","Health":null,"HealthRollup":null},"SupportedEthernetCapabilities":[],"SupportedEthernetCapabilities@odata.count":0,"SupportedLinkCapabilities":[{"AutoSpeedNegotiation":null,"LinkNetworkTechnology":"Ethernet","LinkSpeedMbps":0},{"AutoSpeedNegotiation":null,"LinkNetworkTechnology":"Ethernet","LinkSpeedMbps":0}],"SupportedLinkCapabilities@odata.count":2,"VendorId":"8086","WakeOnLANEnabled":null}

mrlhansen commented 4 weeks ago

The issue is that both ports are reported with NIC.Slot.2 as id, but they should actually be reported as NIC.Slot.2-1 and NIC.Slot.2-2 - which is also the case for my machines (which are running a newer version of both iDRAC and the UEFI). I can implement a fix for this particular edge case, such that the error is not reported. However, I can also see that no relevant information seemingly is being reported (no link status, health status, or link speed). So even if I implement the fix, there will be no useful information. This might change with a firmware update.

love6875520 commented 4 weeks ago

Thanks a lot！

mrlhansen commented 3 weeks ago

I made a new release that includes this fix.

mrlhansen / idrac_exporter

When i enable network metric,i have received this error log #92