Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Navigation  Back up to About Overview 

Known Issues

This section lists the known issues in Network Director Release 3.1.

For the most complete and latest information about known defects Junos Space Network Director defects, use the Juniper Networks online Junos Problem Report Search application.

  • After you upgrade Network Director from Release 2.5 to Release 3.0, the VMware vRealize Operations (vROps) stops receiving data from Network Director.

    Workaround: After upgrading to Network Director Release 3.0, open the vROps user interface and delete the ND adapter, add the ND adapter afresh, and try again. [PR1232254]

  • If you open Network Director in multiple tabs of the same browser window, autorefresh works only for the latest tab.

    Workaround: For autorefresh to work on all sessions, use different browsers. [PR978137]

  • Network Director promotes duplicate VLAN profiles to VLAN with VRRP profiles when you make out-of-band changes to a device.

    Workaround: Delete the unassigned profile from Network Director. [PR1120850]

  • With Microsoft Internet Explorer 11, the View Virtual Network Connectivity page in the Datacenter View might not display the Zoom In, Zoom Out, and Refresh icons.

    Workaround: Scroll to the left of the screen to use the Zoom In, Zoom Out, and Refresh icons. [PR1098365]

  • When you create a Zero Touch Provisioning (ZTP) profile, you are unable to add a default route or gateway address.

    Workaround:

    1. Open the dhcpd.conf file on the DHCP server.

    2. Add an entry for option routers, between the subnet and option subnet-mask entries of the appropriate subnet, and specify the gateway IP address as shown in the example below:

      An example code snippet:

      subnet 10.222.210.0 netmask 255.255.254.0{
      option routers 10.222.210.1;
      option subnet-mask 255.255.254.0;
    3. Restart the dhcpd service from the Junos Space console by issuing the service dhcpd restart command.

      [PR1103325]

  • After you make out-of-band changes on authentication details that are part of a VRRP profile, refreshing the configuration does not create a new profile.

    Workaround: Make the out-of-band changes as part of the VRRP profile, deploy the VRRP profile, and resynchronize the configuration from Network Director. [PR1100717]

  • All the saved changes (not deployed) on the Manage IP connectivity page are lost after you add a new device to the data center.

    Workaround: Before adding a new device to the data center, deploy all the changes that you made in the Manage IP Connectivity page. [PR1109046]

  • After you deploy a policy from Network Director, if you clear the policy options and remove the corresponding Route Filters, only the association of the policy with BGP is deleted. The policy as such is not deleted from the device.

    Workaround: None available at present. [PR1109827]

  • If you have VRRP version 3, after you edit an already deployed Device Common Settings profile and deploy the changes, the deployment fails. This occurs because Network Director appends the VRRP version 3 configuration also to the changed configuration, whereas only the changed configuration must be deployed.

    Workaround: Disable VRRP version 3 or make VRRP configuration compatible with version 3. [PR1118766]

  • In the Manage IP Connectivity workflow, if you opted not to resynchronize devices that are Out of Sync, then all new Save or Deploy operations might fail with the following message Manage IP Connectivity Fails.

    Workaround: Synchronize the devices once prompted. [PR1119978]

  • Deleting configuration files in Junos Space for devices also removes the baseline for the devices.

    Workaround: None available at present. [PR1107121]

  • If an image upgrade job and a cluster fail job runs simultaneously, the image upgrade job status might show as Success, but the image might not have been upgraded.

    Workaround: Retry the image upgrade after cluster failover is complete. [PR1109072]

  • Device might continue to remain in the Synchronising state after some deployments.

    Workaround: Initiate a manual resynchronization for the devices using the Resynchronize Device Configuration task. [PR1111410]

  • During ZTP, device autodiscovery might fail if the device takes long time to reboot after upgrading to the new image.

    Workaround: Discover the device manually if autodiscovery failed after ZTP. [PR1114186]

  • Conversion of two standalone satellite devices to a three member cluster might fail causing the status of one member to be displayed as provision down.

    Workaround: Execute the following command by using the CLI of the aggregation device:run request chassis satellite reboot fpc-slot fpc-slot-number. [PR1200371]

  • The EX9200 device is not identified as an aggregation device in Junos Fusion setup and the link up job is not triggered.

    Workaround: Configure SNMP on EX9200 devices to identify it as an aggregation device. [PR1186656]

  • If two MC-LAG peer devices are brown fielded with different device models and if the IP address of any one of the MC-LAG devices matches the IP address of a Peer device of another MC-LAG device, then the Manage MC-LAG page might display MC-LAG pairing with a different model.

    Workaround: Ensure that MC-LAG peer device has a unique ICCP local IP address in the network. [PR1116833]

  • In the Dashboard View, the data for the last one hour is displayed in grey color after the filter criteria is applied for one hour.

    Workaround: None available at present. [PR1058767]

  • Network Director does not support IPv6 topology discovery.

    Workaround: None available at present. [PR1204972]

  • Network Director might not be able to discover virtual machine tags for vCenter Version 6.0, if there is no description for tags or categories.

    Workaround: Upgrade to vCenter Version 6.0 Update 1 or later and try again. [PR1186865]

  • Brownfield fails if both the apply-groups and the interface description CLI are present under the interface-range stanza. This type of configuration is mostly seen for QFabric CPE.

    Workaround: Delete the apply-groups configuration under the interface-range stanza and specify the groups configuration directly under the interface-range stanza.

    For example in the QFabric CPE configuration, do the following:

    1. Delete the apply-groups configuration under the interface ranges Node_Device_Interfaces and Interconnect_Device_Interfaces:

      user@device#delete interfaces interface-range Node_Device_Interfaces apply-groups qfabric-int
      user@device#delete interfaces interface-range Interconnect_Device_Interfaces apply-groups qfabric-int
    2. Specify the configuration under the groups qfabric-int directly within the interface-range stanza:

      user@device#set interfaces interface-range Node_Device_Interfaces mtu 9216
      user@device#set interfaces interface-range Node_Device_Interfaces unit 0 family ethernet-switching port-mode access vlan members qfabric
      user@device#set interfaces interface-range Interconnect_Device_Interfaces mtu 9216
      user@device#set interfaces interface-range Interconnect_Device_Interfaces unit 0 family ethernet-switching port-mode access vlan members qfabric
    3. Commit the changes:

      user@device#commit

      [PR1125950]

  • In the Datacenter Connectivity View, the zoom pane that appears at the bottom right might not highlight the area that is zoomed.

    Workaround: None available at present. [PR1100350]

  • If you physically connect a new Virtual Chassis leaf member before adding the leaf member from Network Director, the leaf member might not be mapped to Layer 3 Fabrics as expected.

    Workaround: Plug and play is not supported for Virtual Chassis leaf members. Therefore, before you physically connect the Virtual Chassis members, make sure that you add the Virtual Chassis leaf members by using the Layer 3 Fabric wizard. [PR1098910]

  • While editing Layer 3 Fabric profiles (Port, VLAN, and Device profiles) users are not warned by any notification that changing Layer 3 Fabric profile configurations might impact the Layer 3 Fabric functionality.

    Workaround: After the profiles are modified, you can discard the changes from the Deploy task. [PR1058811]

  • For Layer 3 Fabrics, if a plug-and-play leaf device is added and mapped to the Fabric, the Cabling page in the Edit Layer 3 Fabric workflow might not update the cabling plan for that leaf in the graph and grid views.

    Workaround: When you edit the Layer 3 Fabric, modify the description in the Fabric Requirements page and then navigate to the Cabling page. The Cabling page updates the proper cabling plan for the plug-and-play leaf device. [PR1058827]

  • In the manual mode, when out-of-band changes are resynchronized, the conflicting CR might not be listed for quick templates.

    Workaround: None available at present. [PR1046833]

  • In the Edit Layer 3 Fabric workflow, the deployment status might display as Failed for replaced or deleted QFX5100 or EX4300 switches (running Junos OS Release 13.2X51-Dxx) even if the configuration is successfully deployed on the device. Also, it might take more than 10 minutes for the device status to be updated in the Edit Layer 3 Fabric job.

    Workaround: None available at present. [PR1061773]

  • When you create the first data center in Network Director, the View pane might not be updated dynamically with the data center components and devices.

    Workaround: Navigate to any other View and then return to the Datacenter View. [PR1061337]

  • Discovery of a QFabric device that has some system log messages in the default-log-messages file might trigger multiple resynchronization jobs.

    Workaround: Clear the system log messages in the default-log-messages file before discovery. The CLI command for clearing default log messages is clear log default-log-messages. [PR1027051]

  • Network Director does not perform Virtual Chassis resynchronization for standalone devices when the device status changes from DOWN to UP or when Network Director restarts.

    Workaround: Delete and rediscover the devices from Network Director. [PR1001626

  • In Topology View, Network Director does not display LAG details between QFabric and other devices.

    Workaround: None available at present. [PR967224]

  • For wired devices, you are unable to specify the SNMP community string from Network Director. However, the Refresh Discovery task from Topology View requires the SNMP community string to be configured on the device to proceed with the refresh discovery task.

    Workaround: Use the CLI to configure the required community string on the device before you use the Refresh Discovery task. [PR1014926]

  • If you associate a hierarchical Data Center switching or Campus switching ELS CoS profile that has Priority Flow Control configuration with an interface of an EX4300 switch that is part of an EX4600 Virtual Chassis, QFX Virtual Chassis, or Virtual Chassis Fabric, the deployment fails.

    Workaround: Clone the CoS profile, remove the PFC settings from the cloned profile, and associate the cloned profile with the EX4300 interfaces. [PR1017364]

  • When a device is removed from one Virtual Chassis or Virtual Chassis Fabric and added to another Virtual Chassis or Virtual Chassis Fabric respectively, the show virtual-chassis status output command from Network Director might not display the expected information for that member in either the Virtual Chassis or the Virtual Chassis Fabric.

    Workaround: Before you discover that member from Network Director, recycle that member in the Virtual Chassis or Virtual Chassis Fabric in which the member device information is not displayed. [PR970798]

  • If the IP address of space nodes change after installing Network Director, the Layer 3 Fabric configuration might still retain the old IP address of space nodes as the SNMP trap target.

    Workaround: Log in to Junos Space console and select the (Debug) run shell option. Run the following commands at the shell prompt:

    • python2.7 /opt/configure_for_ND.py -ndvip <VIP> -nodeip <node management ip> -restport 20080 -traptgt <node management ip> -dbuser <DB user> -dbpass <DB password> -ndtrapport 10162

    • If there is a device management interface configured, run the same command using device management ip option as follows:
      python2.7 /opt/configure_for_ND.py -ndvip <VIP> -nodeip <device management ip> -restport 20080 -traptgt <device management ip> -dbuser <DB user> -dbpass <DB password> -ndtrapport 10162

    • If it is a cluster environment, the command needs to be run on each node. The argument -traptgt must be repeated the same number of times as the number of nodes present. Assuming there are four nodes, run the following command on each of the four nodes:
      python2.7 /opt/configure_for_ND.py -ndvip <VIP> -nodeip <node management ip> -restport 20080 -traptgt <node1 management ip> -traptgt <node2 management ip> -traptgt <node3 management ip> -traptgt <node4 management ip> -dbuser <DB user> -dbpass <DB password> -ndtrapport 10162

    • If it is a cluster environment and each node has a device management interface configured, the command needs to be run on each node. The argument -traptgt must be repeated the same number of times as the number of nodes present. Assuming there are four nodes, run the following command on each of the four nodes.
      python2.7 /opt/configure_for_ND.py -ndvip <VIP> -nodeip <device management ip> -restport 20080 -traptgt <node1 device management ip> -traptgt <node2 device management ip> -traptgt <node3 device management ip> -traptgt <node4 device management ip> -dbuser <DB user> -dbpass <DB password> -ndtrapport 10162 [PR1062066]

  • Unable to close Details window under Monitor mode.

    Workaround: Press Esc to close the Details window. [PR1117428]

  • Monitor widgets such as Status, Resource Utilization, Power, and Fan details display no data when a few devices are deleted and re-discovered from Network Director.

    Workaround: None available at present. [PR1122181]

  • The Client Sessions and Session Trend widgets might not show any data when a Virtual Chassis member is assigned under Location View or Custom Group View.

    Workaround: Assign the Virtual Chassis logical device as a whole under Location View or Custom Group View. [PR1012400]

  • Running a VM to VM flow analysis might fail or show latency values as 0/0/0 if there are other flows on the VMs being analyzed.

    Workaround: None available at present. [PR1062477]

  • When a QFabric node device alias is changed, Historical Monitoring data will be lost and port entries might be duplicated.

    Workaround: Delete QFabric from Network Director and rediscover the node devices. [PR926912]

  • Network Director does not display alarms at the member or node level for a Virtual Chassis or a Virtual Chassis Fabric.

    Workaround: Select the Virtual Chassis or the Virtual Chassis Fabric device in Logical View to view the alarms. [PR991298]

  • After you upgrade Network Director from Release 2.5 to Release 3.0, you might not be able to add QFX-5200 and QFX5100-24Q switches as leaf devices under Layer 3 Fabrics.

    Workaround: None available at present. [PR1235922]

  • 100-Gigabit Ethernet ports are shown as 40-Gigabit Ethernet ports in the Convert Ports page unless explicitly set in device.

    Workaround: Set the ports with 100-Gigabit transceivers in the device by using the command: set chassis fpc 0 pic 0 port 8 speed 100G and commit the configuration. Wait for the device to be in sync. The Convert Ports page now shows the port correctly as a 100-Gigabit Ethernet port. [PR1234853]

  • Network Director deletes the configuration for ports for which auto-speed detection is disabled in Port Conversion.

    Workaround: None available at present. [PR1228275]

  • The Deploy Configuration page lists all devices for Port Conversion.

    Workaround: Use the Device Selection page to only those devices for which you want to convert ports. [PR1227932]

  • Port Conversion pushes the configuration to incorrect FPCs if the renumbering member ID command is used in the configuration for devices that support VC and VCF configurations.

    Workaround: None available at present. [PR1227497]

  • Network Director might not validate port groups in Port Conversion for QFX10002 and QFX10008 switches.

    Workaround: None available at present. [PR1228276

  • After you upgrade Network Director from Release 2.5 to Release 3.0, incorrect roles are displayed for some of the users.

    Workaround: Roles and tasks are not deleted after the upgrade and only additional tasks are added. Delete these additional tasks after the upgrade. [PR1231329]

  • Unable to create Layer 3 fabrics as Zero Touch Provisioning (ZTP) is not supported on Centos 7.2.1511 and DHCP 4.2.5.

    Workaround: Use the following versions for ZTP support:

    • Ubuntu 14.04

    • Centos 6.6

    [PR1235928]

  • VLANs that are not configured with Layer 3 interfaces are mapped to spine devices.

    Workaround: None available at present. [PR1237441]

  • Device LAG ports configuration is not supported in EVPN-VXLAN.

    Workaround: Use the multi-home LAG option to configure device LAG ports. [PR1236498]

  • For IP Fabrics you cannot configure network address as start address.

    Workaround: Use a valid network IP address in the subnet. [PR1233532]

  • Editing a tenant and assigning ports to a new leaf device, results in failed deployment.

    Workaround: Create a new IP fabric device and assign ports to the device before you edit a tenant and deploy it. [PR1232167

  • The Auto assigned option is not deleted if you remove a Port profile association that is assigned by auto policy and manually assign the same option.

    Workaround: None available at present. [PR1202439]

  • After you delete a port association from Port profile, the LLDP does not show which device is connected when you issue the command delete interface interface-name.

    Workaround: Assign a Port profile to a port manually or by using quick template, to configure unit 0 on the interface. [PR1223305]

  • After adding a cascade port, the satellite cluster is not added to the aggregation devices when the link up event is triggered.

    Workaround:

    1. Connect the device directly to a cascade port on the aggregation device.
    2. Configure the port as a cascade port.
    3. Refresh the topology and verify the configuration generated on the device.
    4. After the resync fusion inventory job is triggered, check the tree to see whether the cluster is added to the Enterprise Fusion setup.

    [PR1228613]

  • If you add more than one port to a Port profile that is associated to the extended Port profiles, the deployment fails.

    Workaround: After the deployment fails, edit the prompted Port profile and deploy the changes again. [PR1237763]

  • The tree view is not correctly displayed when there is more than one cluster with the same name in the tree view.

    Workaround:

    1. Make sure only ICCP and ICL devices are connected, and refresh the topology.

      The link up event is triggered only for redundancy satellite devices, and not for FPC satellite devices.

    2. Connect single-home satellite device or a cluster devices to one of the aggregation device in a multi-host setup and refresh the topology.

      The link up event generates the configuration for different FPC devices.

    [PR1225525]

  • Monitoring cannot be disabled for aggregation devices at the group level.

    Workaround: None available at present. [PR1213805]

  • The configuration for orchestration for ELS devices fails if you select the interface mode as access or if you do not configure an interface mode.

    Workaround: Log in to the device and set interface mode as trunk for the interface and then configure orchestration for that interface. [PR1235223]

  • The Port Conversion page displays for EX4600 switches even though the port conversion is not supported for these devices.

    Workaround: Do not select EX4600 devices for port conversion. [PR1226676]

  • Adding leaf to an existing multihome LAG, editing the multihome ae inetrface, or deleting an existing leaf from the LAG are not supported in an overlay network.

    Workaround: None available at present. [PR1206683]

  • After an upgrade to Network Director 3.0, earlier versions of APIs are listed.

    Workaround: Right-click on the APIs and select Uninstall to remove the earlier versions. [PR1229453]

  • The quick template functionality is not supported in Junos Fusion Enterprise devices.

    Workaround: None available at present. [PR1239050]

  • In a medium-scale Junos Fusion setup where five Junos Fusion systems each with 100 Satellite Devices are managed, incorrect monitoring data is displayed.

    Workaround: In a Junos Fusion setup where 5 Junos Fusion systems each with 100 Satellite Devices are managed, change the number of parallel requests by setting the system property to 5 from default value 25.

    To change the parallel requests value:

    1. Open a debug (command) prompt by using the Junos Space Settings Menu.
    2. Navigate to /usr/local/jboss/domain/configuration/host.xml
    3. Under <jvm-options>, change the option value <option value="-Dmonitoring.simultaneous.request=5"/>
    4. Restart the jboss process.

      service jboss restart

    [PR1239748]

  • When large number of devices are discovered in Network Director, some of the device status is shown as Down in Network Director even though their status is shown as UP in Junos Space Management Platform.

    Workaround: Delete devices that are incorrectly shown as Down in Network director and re-discover the devices in Network Director. [PR1207914]

  • In large-scale environments, some of the devices cannot be monitored after the device discovery.

    Workaround: Restart the JBoss server for the monitoring features to work properly in standalone and cluster setups:

    To restart the JBoss server in a standalone setup:

    1. Stop the watchdog, domain controller, and JBoss services on the standalone node.

      service jmp-watchdog stop

      service jboss-dc stop

      service jboss stop

    2. Start the watchdog service.

      service jmp-watchdog start

      Note: Starting the watchdog service restarts the JBoss and domain controller services as well.

    To restart the JBoss server in a cluster setup:

    1. Stop the services on the secondary node.

      service jmp-watchdog stop

      service jboss stop

    2. Stop the services on the master node (You can find the VIP hosted node Space > Fabric)

      service jmp-watchdog

      stop service jboss-dc

      stop service jboss stop

    3. Start the services on the master node.

      service jmp-watchdog start

    4. Start the service on the secondary node.

      service jmp-watchdog start

    [PR1165010]

  • The Port status and the power supply status is not shown in a cluster setup when two satellite devices are configured as member nodes.

    Workaround: None available at present. [PR1223602]

  • The overlay fabric configuration does not get deleted when you remove an IP fabric from the Selected section of the Edit Overlay Fabric page. However, the overlay fabric configuration gets deleted if the overlay is deleted from the from the Manage Overlay Fabric page.

    Workaround: None available at present. [PR1234916]

  • ESXi server does not work on ports that have only one VLAN configured. When a tenant overlay fabric is configured such that one of the ports is configured with only one VLAN, native VLAN ID is set to the ID of the only VLAN configured on the interface. This allows Bare Metal Servers (BMS) to use untagged packets. For an ESXi server to work on such ports, the native VLAN ID needs to be removed.

    Workaround: Deploy a template by using the command delete interface native-vlan-id for those interfaces where one VLAN is configured and tagged packets are sent. ESXi server uses tagged packets by default. [PR1234899]

  • Duplicate Port profiles are shown in Network Director after you upgrade Network Director from:

    • Release 2.5R1 to Release 3.0R1 or 3.0R2

    • Release 2.5R2 to Release 3.0R1 or 3.0R2

    Workaround: Delete all the unassigned profiles that are generated as duplicate Port profiles. You can identify duplicate Port profiles from the profile name. A duplicate Port profile name is suffixed with a number. For example, iSCSI_918. [PR1246846]

  • High CPU utilization is observed in the EX2200, EX2300, EX3400, or EX3300 devices when these devices are managed and monitored in Network Director.

    Workaround: Increase the poll interval from default 10 mins to 40 or 50 mins for the categories PortTrafficMonitorCollector and EquipmentMonitorDeviceStatusCollector under monitor settings.

    To increase the poll interval:

    1. Navigate to Preferences > Monitoring > Monitoring Settings.

      The Preferences page displays the monitoring settings.

    2. In the Poll Interval (in mins) field, enter the poll interval for the categories PortTrafficMonitorCollector and EquipmentMonitorDeviceStatusCollector.
    3. Click OK.

    Note: The poll interval for both the categories should be the same.

    If increasing poll interval does not reduce the CPU usage, disable monitoring for affected devices.

    To disable monitoring for the devices:

    1. Navigate to Preferences > Monitoring > Device Settings.

      The Preferences page displays Enable/Disable Collection for devices in a tabular format.

    2. From the table, select the check box that corresponds to the device for which you want to disable the monitoring.
    3. Click Disable.
    4. Click OK.

    [PR1285504]

  • In a Junos Fusion Data Center setup, if the alias name of the satellite device does not match on both the aggregation devices, Network Director does not show the monitoring data for the satellite devices.

    Workaround: Configure the same alias name for the satellite device on both the aggregation devices. [PR1293816 ]

  • After the IP Fabric provisioning is successfully completed, auto discovery of IP fabric is not getting triggered in Network Director. This is observed when the eth3 interface IP address is changed on Junos Space sever after Junos Space installation.

    Workaround: Modify the eth3 IP address in the .conf files that are pushed to the TFTP server when provisioning the IP Fabric. This ensures that new eth3 interface IP address is used by Network Director while discovering the IP Fabric. [PR1293526]

  • In Network Director Wired Device Common Settings, you cannot disable the DHCP relay on a device.

    Workaround: To disable the DHCP relay on a device, access the device by using the CLI and remove the DHCP relay config stanza from the device. [PR1290375]

  • In Network Director Wired Device Common Settings, you cannot delete the profile login settings relay on a device.

    Workaround: To delete the profile login from a device, access the device by using the CLI and remove the Profile Login configuration stanza from the device. [PR1291978]

  • When the Connection Association Key and Confirm Connection Association Key are edited in a MACsec profile, the MACsec profile does not show the state as Pending Deployment.

    Workaround: Edit some other attribute in the MACsec profile along with the Connection Association Key and Confirm Connection Association Key for the MACsec profile to show the state as Pending Deployment. [PR1295679]

Modified: 2018-04-11