Devices Window

The Devices window shows data pertaining to the physical devices in a tabular format.

Devices Window Data

Data Type	Description
Health	Health of the device reflecting the highest alarm severity. Please refer to the Health States table.
Name	Name of the device Note If UFM Agent is running on a device, the following icon will appear next to the device name:
GUID	System GUID of the device
Type	Type of the device: switch, node, IB router, and getaway
IP	IP address of the device
Vendor	The vendor of the device
Firmware Version	The firmware version installed on the device

Health States

Icon	Name	Description
	Normal	Information/notification displayed during normal operating state or a normal system event.
	Critical	Critical means that the operation of the system or a system component fails.
	Minor	Minor reflects a problem in the fabric with no failure.
	Warning	Warning reflects a low priority problem in the fabric with no failure. A warning is asserted when an event exceeds a predefined threshold.

A right-click on the device name displays a list of actions that can be performed on it.

Devices Actions

This action opens a terminal window that connects to the UFM server via WebSocket, which then establishes an SSH connection to the switch using a TCP socket.

Start by selecting the "Open Terminal" option:

A terminal window will then open.

Clicking the minimize button will put the session into minimized mode, allowing it to remain open.

This action opens a new browser tab that shows the switch's web interface.

First, select "Open Webpage."

A new page displaying the switch's web interface will then appear in your browser.

From the Devices table, it is possible to mark devices as healthy or unhealthy using the context menu (right-click).

There are two options for marking a device as unhealthy:

• Isolate

• No Discover

Server: conf/opensm/opensm-health-policy.conf content:

            

            
0xe41d2d030003e3b0 34 UNHEALTHY isolate
0xe41d2d030003e3b0 19 UNHEALTHY isolate
0xe41d2d030003e3b0 3 UNHEALTHY isolate
0xe41d2d030003e3b0 26 UNHEALTHY isolate
0xe41d2d030003e3b0 0 UNHEALTHY isolate
0xe41d2d030003e3b0 27 UNHEALTHY isolate
0xe41d2d030003e3b0 7 UNHEALTHY isolate
0xe41d2d030003e3b0 10 UNHEALTHY isolate
0xe41d2d030003e3b0 11 UNHEALTHY isolate
0xe41d2d030003e3b0 22 UNHEALTHY isolate
0xe41d2d030003e3b0 18 UNHEALTHY isolate
0xe41d2d030003e3b0 29 UNHEALTHY isolate
0xe41d2d030003e3b0 8 UNHEALTHY isolate
0xe41d2d030003e3b0 5 UNHEALTHY isolate
0xe41d2d030003e3b0 17 UNHEALTHY isolate
0xe41d2d030003e3b0 23 UNHEALTHY isolate
0xe41d2d030003e3b0 15 UNHEALTHY isolate
0xe41d2d030003e3b0 24 UNHEALTHY isolate
0xe41d2d030003e3b0 2 UNHEALTHY isolate
0xe41d2d030003e3b0 16 UNHEALTHY isolate
0xe41d2d030003e3b0 13 UNHEALTHY isolate
0xe41d2d030003e3b0 14 UNHEALTHY isolate
0xe41d2d030003e3b0 32 UNHEALTHY isolate
0xe41d2d030003e3b0 33 UNHEALTHY isolate
0xe41d2d030003e3b0 35 UNHEALTHY isolate
0xe41d2d030003e3b0 20 UNHEALTHY isolate
0xe41d2d030003e3b0 21 UNHEALTHY isolate
0xe41d2d030003e3b0 28 UNHEALTHY isolate
0xe41d2d030003e3b0 1 UNHEALTHY isolate
0xe41d2d030003e3b0 9 UNHEALTHY isolate
0xe41d2d030003e3b0 4 UNHEALTHY isolate
0xe41d2d030003e3b0 31 UNHEALTHY isolate
0xe41d2d030003e3b0 30 UNHEALTHY isolate
0xe41d2d030003e3b0 36 UNHEALTHY isolate
0xe41d2d030003e3b0 12 UNHEALTHY isolate
0xe41d2d030003e3b0 25 UNHEALTHY isolate
0xe41d2d030003e3b0 6 UNHEALTHY isolate
        

/opt/ufm/files/log/opensm-unhealthy-ports.dump content:

Server /opt/ufm/files/conf/opensm/opensm-health-policy.conf content:

            

            
0xe41d2d030003e3b0 15 HEALTHY
0xe41d2d030003e3b0 25 HEALTHY
0xe41d2d030003e3b0 35 HEALTHY
0xe41d2d030003e3b0 0 HEALTHY
0xe41d2d030003e3b0 11 HEALTHY
0xe41d2d030003e3b0 21 HEALTHY
0xe41d2d030003e3b0 28 HEALTHY
0xe41d2d030003e3b0 7 HEALTHY
0xe41d2d030003e3b0 17 HEALTHY
0xe41d2d030003e3b0 14 HEALTHY
0xe41d2d030003e3b0 24 HEALTHY
0xe41d2d030003e3b0 34 HEALTHY
0xe41d2d030003e3b0 3 HEALTHY
0xe41d2d030003e3b0 10 HEALTHY
0xe41d2d030003e3b0 20 HEALTHY
0xe41d2d030003e3b0 31 HEALTHY
0xe41d2d030003e3b0 6 HEALTHY
0xe41d2d030003e3b0 16 HEALTHY
0xe41d2d030003e3b0 27 HEALTHY
0xe41d2d030003e3b0 2 HEALTHY
0xe41d2d030003e3b0 13 HEALTHY
0xe41d2d030003e3b0 23 HEALTHY
0xe41d2d030003e3b0 33 HEALTHY
0xe41d2d030003e3b0 30 HEALTHY
0xe41d2d030003e3b0 9 HEALTHY
0xe41d2d030003e3b0 19 HEALTHY
0xe41d2d030003e3b0 26 HEALTHY
0xe41d2d030003e3b0 36 HEALTHY
0xe41d2d030003e3b0 5 HEALTHY
0xe41d2d030003e3b0 12 HEALTHY
0xe41d2d030003e3b0 22 HEALTHY
0xe41d2d030003e3b0 32 HEALTHY
0xe41d2d030003e3b0 1 HEALTHY
0xe41d2d030003e3b0 8 HEALTHY
0xe41d2d030003e3b0 18 HEALTHY
0xe41d2d030003e3b0 29 HEALTHY
0xe41d2d030003e3b0 4 HEALTHY
        

/opt/ufm/files/log/opensm-unhealthy-ports.dump content:

            

            
# NodeGUID, PortNum, NodeDesc, PeerNodeGUID, PeerPortNum, PeerNodeDesc, {BadCond1, BadCond2, ...}, timestamp
        

Software/Firmware Upgrade via FTP

Software and firmware upgrade over FTP is enabled by the UFM Agent. UFM invokes the Software/Firmware Upgrade procedure locally on switches or on hosts. The procedure copies the new software/firmware file from the defined storage location and performs the operation on the device. UFM sends the set of attributes required for performing the software/firmware upgrade to the agent.

The attributes are:

• File Transfer Protocol – default FTP

  • The Software/Firmware upgrade on InfiniScale III ASIC-based switches supports FTP protocol for transmitting files to the local machine.

  • The Software/Firmware upgrade on InfiniScale IV-based switches and hosts supports TFTP and protocols for transmitting files to the local machine.

• IP address of file-storage server

• Path to the software/firmware image location

  The software/firmware image files should be placed according to the required structure under the defined image storage location. Please refer to section Devices Window.

• File-storage server access credentials (User/Password)

In-Band Firmware Upgrade

You can perform in-band firmware upgrades for externally managed switches, which requires recognition of the current PSID. For more information, please refer to the PSID and Firmware Version In-Band Discovery. UFM uses flint, a firmware burning tool, from the MFT for the in-band firmware update process.

Before proceeding with the upgrade, ensure that you create the firmware repository on the UFM server in the directory /opt/ufm/files/userdata/fw/. A separate subdirectory should be created for each PSID, with the corresponding firmware image placed under it. For example:

            

            
/opt/ufm/files/userdata/fw/
	 MT_0D80110009
			fw-ConnectX2-rel-2_9_1000-MHQH29B-XTR_A1.bin
	 MT_0F90110002
			fw-IS4-rel-7_4_2040-MIS5023Q_A1-A5.bin
        

Directory Structure for Software or Firmware Upgrade Over FTP

Before performing a software or firmware upgrade, you must create the following directory structure for the upgrade image. The path to the <ftp user home>/<path>/ directory should be specified in the upgrade dialog box.

            

            
<ftp user home>/<path>/
	 InfiniScale3 - For anafa based switches Software/Firmware upgrade images
			voltaire_fw_images.tar – firmware image file
			ibswmpr-<s/w version>.tar – software image file
	InfiniScale4 - For InfiniScale IV based switches Software/Firmware upgrade images
			firmware_2036_4036.tar – Firmware image file
			upgrade_2036_4036.tgz – Software image file
	OFED /* For host SW upgrade*/
			OFED-<OS label>.tar.bz2
	<PSID>* – For host FW upgrade
			fw_update.img
        

The <PSID> value is extracted from the mstflint command:

            

            
mstflint -d <device> q
        

The device is extracted from the lspci command. For example:

            

            
# lspci
06:00.0 InfiniBand: Mellanox Technologies MT25208 InfiniHost III Ex
# mstflint -d 06:00.0 q | grep PSID
PSID: VLT0040010001
        

PSID and Firmware Version In-Band Discovery

The device PSID and device firmware version are required for in-band firmware upgrade and for the correct functioning of Subnet Manager plugins, such as Congestion Control Manager and Lossy Configuration Management. For most devices, UFM discovers this information and displays it in the Device Properties pane. The PSID and the firmware version are discovered by the Vendor-specific MAD.

By default, the gv.cfg file value for event_plugin_option is set to (null). This means that the plugin is disabled and opensm does not send MADs to discover devices' PSID and FW version. Therefore, values for devices' PSID and FW version are taken from ibdiagnetoutput (section NODES_INFO).

The below is an example of the default value:

            

            
event_plugin_options = (null)
        

To enable the vendor-specific discovery by OpenSM, in the gv.cfg configuration file, change the value of event_plugin_option to (--vendinfo –m 1), as shown below:

            

            
event_plugin_options = --vendinfo –m 1
        

If the value is set to –vendinfo –m 1, the data should be supplied by opensm, and in this case the ibdiagnetoutput is ignored.

Switch Management IP Address Discovery

From NVIDIA switch FM version 27.2010.3942 and up, NVIDIA switches support switch management IP address discovery using MADs. This information can be retrieved as part of ibdiagnetrun (ibdiagnetoutput), and assigned to discover switches in UFM.

There is an option to choose the IP address of which IP protocol version that is assigned to the switch: IPv4 or IPv6.

The discovered_switch_ip_protocolkey, located in the gv.cfg file in section [FabricAnalysys], is set to 4 by default. This means that the IP address of type IPv4 is assigned to the switch as its management IP address. In case this value is set to 6, the IP address of type IPv6 is assigned to the switch as its management IP address.

After changing the discover_switch_ip_protocol value in gv.cfg, the UFM Main Model needs to be restarted for the update to take effect. The discovered IP addresses for switches are not persistent in UFM – every UFM Main Model restarts the values of management IP address which is assigned from the ibdiagnetoutput.

Upgrading Server Software

The ability to update the server software is applicable only for hosts (servers) with the UFM Agent.

To upgrade the software:

1. Select a device.

2. From the right-click menu, select Software Update.

3. Enter the parameters listed in the following table.

   Parameter	Description
   Protocol	Update is performed via FTP protocol
   IP	Enter the host IP
   Path	Enter the parent directory of the FTP directory structure for the Upgrade image. The path should not be an absolute path and should not contain the first slash (/) or trailer slash.
   User	Name of the host username
   Password	Enter the host password

4. Click Submit to save your changes.

Upgrading Firmware

You can upgrade firmware over FTP for hosts and switches that are running the UFM Agent, or you can perform an in-band upgrade for externally managed switches and HCAs.

Before you begin the upgrade ensure that the new firmware version is in the correct location. For more information, please refer to section In-Band Firmware Upgrade.

To upgrade the firmware:

1. Select a host or server.

2. From the right-click menu, select Firmware Upgrade.

3. Select protocol In Band.

4. For upgrade over FTP, enter the parameters listed in the following table.

   Parameter	Description
   IP	Enter device IP
   Path	Enter the parent directory of the FTP directory structure for the Upgrade image. The path should not be an absolute path and should not contain the first slash (/) or trailer slash.
   Username	Name of the host username
   Password	Enter the host password

5. Click submit to save your changes.

   Note

   The firmware upgrade takes effect only after the host or externally managed switch is restarted.

Upgrade Cables Transceivers Firmware Version

The main purpose of this feature is to add support for burning of multiple cables transceiver types on multiple devices using linkx tool which is part of flint. This needs to be done from both ends of the cable (switch and HCA/switch).

To upgrade cables transceivers FW version:

1. Navigate to managed elements page

2. Select the target switches and click on Upgrade Cable Transceivers option

   

3. A model will be shown containing list of the active firmware versions for the cables of the selected switches, besides the version number, a badge will show the number of matched switches:

   

   

4. After the user clicks Submit, the GUI will start sending the selected binaries with the relevant switches sequentially, and a model with a progress bar will be shown (this model can be minimized):

   

5. After the whole action is completed successfully, you will be able to see the following message at the model bottom The upgrade cable transceivers completed successfully, do you want to activate it? by clicking the yes button it will run a new action on all the burned devices to activate the new uploaded binary image.

6. Another option to activate burned cables transceivers you can go to the Groups page and right click on the predefined Group named Devices Pending FW Transceivers Reset or you can right click on the upgraded device from managed element page and select Activate cable Transceivers action.

   

Selecting a device from the Devices table reveals the Device Information table on the right side of the screen. This table provides information on the device’s ports, cables, groups, events, alarms, , and device access.

General Tab

Provides general information on the selected device.

Ports Tab

This tab provides a list of the ports connected to this device in a tabular format.

Ports Data

Data Type	Description
Port Number	The number of ports on device.
Node	The node name/GUID/IP that the port belongs to. Note that you can choose the node label (name/GUID/IP) using the drop-down menu available above the Ports data table.
Health	Health of the port reflecting the highest alarm severity. Please refer to the Health States table.
State	Indicates whether the port is connected (active or inactive).
LID	The local identifier (LID) of the port.
MTU	Maximum Transmission Unit of the port.
Speed	Lists the highest value of active, enabled and supported speeds in icons indicating their status: Dark green – active speed Light green – enabled speed Grey – supported yet disabled speed
Width	Lists the highest value of active, enabled and supported widths in icons indicating their status: Dark green – active width Light green – enabled width Grey – supported yet disabled width
Peer	The GUID of the device the port is connected to.
Peer Port	The name of the port that is connected to this port.

Cables Tab

This tab provides a list of the cables connected to this device in a tabular format.

Cables Data

Groups Tab

This tab provides a list of the groups to which the selected device belongs.

Groups Data

Alarms Tab

This tab provides a list of all UFM alarms related to the selected device.

Alarms Data

Events Tab

This tab provides a list of the UFM events that are related to the selected device.

Events Data

Inventory Tab

This tab provides a list of the device’s modules with information in a tabular format.

Inventory Data

HCAs Tab

This tab provides a list of the device’s HCAs with information in a tabular format.

Device Access Tab

This tab allows for managing the access credentials of the selected device for remote accessibility. To be able to set access credentials for the device, a device IP must be set either by installing UFM Agent on the device, or by manually setting the IP under IP Address Settings (IP is now supported with v4 and v6).

To edit your device access credentials

1. Select the preferred protocol tab:

   • SSH – allows you to define the SSH parameters to open an SSH session on your device (available for nodes and switches)

   • IPMI – allows you to set the IPMI parameters to open an IPMI session on your device for remote power control (available for nodes only)

   • HTTP – allows you to define the HTTP parameters to open an HTTP session on your device (available for switches only)

2. Click Update to save your changes.

   

Device Access Credentials Parameters

Virtual Networking Tab

This tab displays a map containing the HCAs for the selected device, and the ports and virtual ports it is connected to.