Open Defects

The following defects are open in EFA 2.4.4.

Parent Defect ID: EFA-5592 Issue ID: EFA-5592
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: Certificates need to be manually imported on replaced equipment in-order to perform RMA.
Condition: RMA/replaced equipment will not have ssh key and auth certificate, in-order to replay the configuration on new switch user needs to import the certificates manually.
Workaround:

import certificate manually

efa certificates device install --ips x,y --certType

Parent Defect ID: EFA-5732 Issue ID: EFA-5732
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: When firmware download is in progress, fabric delete command is accepted without an error.
Condition: If fabric delete command is submitted when firmware download is in progress, it fails.
Workaround:

Allow firmware download process to complete. Status of the same can be checked using command

efa inventory device firmware-download show --fabric {fabric name}

Recovery: Fabric can be deleted once the firmware download is completed
Parent Defect ID: EFA-5841 Issue ID: EFA-5841
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: When firmware download is in progress, tenant create command is accepted without an error.
Condition: If tenant commands are submitted when firmware download is in progress, it results in erroneous configuration and some configurations may miss.
Workaround:

Allow firmware download process to complete. Status of the same can be checked using command

efa inventory device firmware-download show --fabric {fabric name}

Recovery: Tenant commands can be submitted after the firmware download is completed
Parent Defect ID: EFA-5874 Issue ID: EFA-5874
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: On device registration, the IP of the EFA system is recorded in the logging entry on the device so logs can be forwarded to the EFA system for notification. When the EFA system is backed up and restored on another system with a different IP, the old IP of the EFA system is still present on the devices and the devices will continue to forward logs to the old EFA IP.
Workaround: Users will have to manually login to each devices and remove the logging entry for the old EFA IP.
Parent Defect ID: EFA-5927 Issue ID: EFA-5927
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: Configuration reconciliation fails with an error "drift and reconcile failed waiting for status from tenant." because of the timeout.
Condition:

When the switch configurations drift from the intended configurations in EFA due to scenarios as follows:

1. L3 Epg is created with large ctag-range (e.g. 2-2000)

2. EFA configured VLANs and PO configurations are manually removed from the switch.

3. Switch is reloaded in maintenance mode

Recovery: After the switch is moved out of maintenance mode after reload, configuration drift can be viewed and reconciled using "efa inventory drift-reconcile execute --reconcile --ip <switch-ip>" CLI.
Parent Defect ID: EFA-5928 Issue ID: EFA-5928
Severity: S2 – High
Product: Extreme Fabric Automation Reported in Release: EFA 2.2.0
Symptom: Configuring devices to default startup-config and adding them to a non-clos fabric does not enable all MCT ports resulting into fabric validation failure for missing link
Condition: Added devices immediately after setting to default startup config
Workaround:

Remove the devices from fabric and re-add

efa fabric device remove --name <fabric-name> --ip <device-ips>

efa inventory device delete --ip <device-ips>

efa fabric device add-bulk --name <fabric-name> --rack <rack-name> --username <username> --password <password> --ip <device-ips>

Parent Defect ID: EFA-6501 Issue ID: EFA-6501
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.3.0
Symptom: Configuration Drift for VRF still shown in "cfg-in-sync" though its child configuration are drifted on SLX switch.
Condition:

With below steps issue can be observed.

1) Create VRF/EPG having route target, static route and bgp configuration.

2) Introduce drift in VRF route target or static route or bgp configuration on SLX switch.

3) Update device from efa command "efa inventory device update --ip <device ip>"

4) Check device drift using efa command as "efa inventory drift-reconcile execute --ip <device ip>"

5) VRF shows as "cfg-in-sync" though its child configuration was drifted.

Workaround: None
Recovery: After drift and reconcile all EFA and device configuration will be in sync.
Parent Defect ID: EFA-7324 Issue ID: EFA-7324
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.3.0
Symptom: Continuous create/delete of BGP peer-group and peer can finally cause CLI errors
Condition: When create and delete BGP peer/peer-group is repeatedly done in a loop. This will cause inventory does not have chance to update its DB with the current design so DB can be out of sync between inventory and tenant. When other events happen such as timer collection from inventory to sweep config to tenant, it can cause issues on tenant DB where CLI can fail.
Workaround: Avoid such cycles of operations
Recovery: Delete the BGP peer/peer-group in problem and recreate them again.
Parent Defect ID: EFA-8090 Issue ID: EFA-8090
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: When a fabric containing more than 15 newly registered devices is deployed using the CLI 'efa fabric configure', an attempt to add ports of any of these devices to a tenant within 5 minutes may fail. The error will indicate that the ports have not yet been registered in the fabric
Condition: Attempt to add device ports of a recently configured fabric to a tenant may fail with an error indication that the ports have not yet been registered in the fabric
Workaround: Wait for up to 5 minutes after deploying the fabric before adding ports to a tenant
Recovery: This is a transient error. Rerunning the port-add operation after a maximum wait time of 5 minutes will succeed
Parent Defect ID: EFA-8152 Issue ID: EFA-8152
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: While graceful-restart(GR) updating with value TRUE and inflight transition triggered as a part of EFA rollover then update will continue as a part of inflight transition.
Condition: Update GR with value TRUE and perform EFA rollover on HA setup.
Parent Defect ID: EFA-8155 Issue ID: EFA-8155
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: "cluster-client auto" is not configured under port channel for first reloaded device.
Condition:

Execute below steps to hit this condition

1) Create fabric on MCT paired device

2) Create Tenant/PO/VRF/EPG

3) Enable MM mode on both device

4) Perform EFA backup

5) Delete EPG/VRF/PO/Tenant

6) Delete fabric

7) Restore EFA backup

8) Reload device one by one

After these steps check PO on both device, "cluster-client auto" will not configured on first reloaded device.

Workaround: Instead of reload device in Step (8), perform manual DRC using inventory CLI as "efa inventory drift-reconcile execute --ip <device ip> --reconcile" for each device.
Parent Defect ID: EFA-8257 Issue ID: EFA-8257
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EFA is not able to detect drift for configuration like VRF/VE/VLAN/EVPN
Condition:

Please follow below steps,

1) Create tenant/VRF/PO/EPG

2) As soon as EPG creation pushed configuration on device, remove them from device.

3) Check drift using inventory CLI as "efa inventory drift-reconcile execute --ip --device-ip <device ip>"

Workaround: As this is timing issue so we need to wait for 1 min before remove configurations from device.
Recovery: We need to delete EPG and recreate it again.
Parent Defect ID: EFA-8269 Issue ID: EFA-8269
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EPG app-state moved to cfg-refresh-err after epg delete and admin up
Condition:

1) Configure clos fabric (Medium scale fabric)

2) Create tenant

3) Admin down the devices

4) Create port-channels, vrfs and epgs

5) Admin up the following devices

Wait for the DRC to be success

6) Repeat step 3

Wait for the devices to put into maintenance mode

7) Create bgp peer-group and dynamic peers

8) Delete all epg's

9) Repeat step 5

10) Vrfs are getting deleted from admin up devices

11) EPG app-state move to cfg-refresh-err

Recovery: Delete the EPGs in cfg-refresh-err state and recreate them.
Parent Defect ID: EFA-8273 Issue ID: EFA-8273
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EPG Update "vrf-add" operation gives success when EPG is in "vrf-delete-pending" state
Condition: Perform EPG Update "vrf-add" operation on an EPG in "vrf-delete-pending" state
Workaround: No workaround
Recovery: User needs to remove the VRF from EPG using EPG update "vrf-delete" operation before attempting the "vrf-add" operation.
Parent Defect ID: EFA-8297 Issue ID: EFA-8297
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom:

EPG update anycast-ip-delete operation succeeded for deletion of provisioned anycast-ip for admin-down device.

This issue is observed only if an update anycast-ip-add operation is performed after device is put in admin down state and the new config is in non-provisioned state followed by anycast-ip-delete operation for already configured anycast-ip.

Condition:

Steps to reproduce issue:

1) Configure EPG with anycast-ip (ipv4/ipv6)

2) Make one device admin-down

3) Anycast-ip update-add new anycast-ip (ipv6/ipv4)

4) Update-delete provisioned anycast-ip configured in step-1 (ipv4/ipv6)

Step (4) should fail as IP is already configured on the device and trying to delete it should fail as part of APS.

Workaround: No workaround for this.
Recovery: Recovery can be done by configuring EPG again with the required configuration using efa or cleaning device config for anycast-ip on the switch.
Parent Defect ID: EFA-8315 Issue ID: EFA-8315
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: User adds ports in empty EPG and immediately deletes them. The following adding ports into EPG can have error as duplicate entry
Condition:

1) Add ports in empty EPG

2) Delete ports from epg right away

3) Add ports into EPG. Which can have error.

Workaround: After adding ports into EPG, wait certain time before trying to delete ports from EPG.
Recovery: Delete the EPG and recreate again.
Parent Defect ID: EFA-8319 Issue ID: EFA-8319
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: PO creation failed with error "Devices are not MCT Pairs".
Condition:

Please follow below steps,

1) Create fabric/tenant/PO/EPG

2) Take EFA backup

3) Delete EPG/PO/tenant/fabric

4) Restore EFA backup taken in step (2)

5) Delete tenant from which was created before backup

6) Create same tenant again

7) Create PO under same tenant

Workaround:

As after restore MCT peer details are Nil so we need to perform DRC after restore taken backup.

After step (4) above, we need to perform DRC using inventory CLI as

efa inventory drift-reconcile execute --ip <device ip 1> --reconcile

efa inventory drift-reconcile execute --ip <device ip 2> --reconcile

Parent Defect ID: EFA-8322 Issue ID: EFA-8322
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EPG Update "anycast-ip-delete" operation gives different output/result when one of the EPG device is admin down
Condition:

1) Create L3 EPG with anycast-ip/anycast-ipv6

2) Take one of EPG device administratively down

3) Bring device admin up which was taken down in previous step

4) While device is coming up administratively, try EPG Update "anycast-ip-delete" operation

Workaround: No workaround
Recovery: No recovery as such. Wait for device to be completely up before trying EPG Update "anycast-ip-delete" operation
Parent Defect ID: EFA-8334 Issue ID: EFA-8334
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: system backup and restore causes epg state to be in cfg-refresh-err
Condition: Tenant DB and inventory DB needs time to be in sync. In a busy and scaled system this in sync can take much longer time to finish. Backup DB during DB un-sync window can cause system saves the DBs for tenant and inventory which are not synced yet and following restore will have issues.
Workaround: If there's a need to make system backup, please execute the backup after system have not made any new config for few minutes. It's needed for the inventory and tenant databases to be in sync before executing system backup. In a busy system the DB sync can take longer to finish.
Recovery: Delete the EPGs which report errors and recreate them.
Parent Defect ID: EFA-8335 Issue ID: EFA-8335
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: system backup and restore causes following manual DRC has errors
Condition: Tenant DB and inventory DB needs time to be in sync. In a busy and scaled system this in sync can take much longer time to finish. Backup DB during this window will cause system saves the DB for tenant and inventory which is not synced yet and following restore will have issues.
Workaround: If there's a need to make system backup, please execute the backup after system have not made any new config for few minutes. It's needed for the inventory and tenant databases to be in sync before executing system backup. In a busy system the DB sync can take longer to finish.
Recovery: Delete the epg or tenant with problem and recreate them.
Parent Defect ID: EFA-8391 Issue ID: EFA-8391
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom:

1) NetConf error for epg update port-group-add operation. Caused due to conflicting anycast-ip added to EPG using anycast-ip-add operation.

2) Error for epg update port-group-add operation. Caused due to configuring different anycast-ip for same vlan in EPG using anycast-ip-add operation.

Condition:

1) If the user provides conflicting anycast-ip in empty EPG then it should throw an error.

* Create EPG1 with port/po, vrf, vlan, anycast-ip

* Create EPG2 without port/po

* Add new VRF to EPG2

* Add conflicting anycast-ip already used in EPG1 with different vrf

* Add port to EPG

2) Multiple EPG sharing the same VRF, VLAN with different anycast-ip should throw an error.

* Create EPG1 with port/po, vrf, vlan, anycast-ip

* Create EPG2 without port/po

* Add vrf to EPG1 which is used in EPG2

* Add new anycast-ip to EPG

* Add port to EPG (This will cause conflict)

Workaround: NA
Recovery:

Delete conflicting anycast-ip from EPG.

Add correct anycast-ip.

Add port/po to EPG.

Parent Defect ID: EFA-8408 Issue ID: EFA-8408
Severity: S4 - Low
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EPG create failure while testing brownfield scenario by creating EPG with the same VRF/SR/SRBfd configuration as present on the device.
Condition:

Creating EPG with the same VRF configuration fails due to static route key mismatch.

Steps:

1) Create Tenant, Vrf (with staticRoute), EPG

2) Check "sh run vrf" and "sh run router bgp" on device

3) delete EPG

4) Create VRF on device directly using config in step 2

5) Update inventory service

6) Try to create EPG with same VRF. (compareVrf fails due to key mismatch)

Workaround: No workaround.
Recovery:

1) Remove Vrf from SLX.

2) Update inventory using "efa inventory device update ..."

3) Create EPG using "efa tenant epg create ..."

Parent Defect ID: EFA-8443 Issue ID: EFA-8443
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: For Tenant created with L3 port having multiple ip-address associated with it, "efa tenant show" will have repeated entries of that L3 port.
Condition:

Steps to reproduce issue:

1) Assign multiple IPs to the physical port on SLX.

2) Create Tenant using same L3 port.

3) Check Tenant show output.

L3 ports having multiple IPs will have repeated entry in the "efa tenant show" output.

Workaround: No workaround.
Recovery: Recovery can be done by removing all but one IP from the L3 port on SLX followed by an inventory device update.
Parent Defect ID: EFA-8448 Issue ID: EFA-8448
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom:

When the ports provided by the user in “tenant update port-delete operation” contains all the ports owned by the port-channel, the PO goes into delete pending state. However, the ports are not deleted from the PO.

They get deleted from the tenant though.

Condition: This issue is seen when the ports provided by the user in “tenant update port-delete operation” contains all the ports owned by the port-channel resulting in an empty PO.
Workaround: User needs to provide ports for “tenant update port-delete operation” which do not result in an empty PO i.e. PO needs to have at least 1 member port.
Recovery: Add the ports back using "tenant port-add operation" so that the port-channel has at least 1 member port. The use "efa configure tenant port-channel" to bring the po back to stable state.
Parent Defect ID: EFA-8465 Issue ID: EFA-8465
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: The "efa inventory device firmware-download prepare add" command fails with "Please specify 'fullinstall' option in firmware download cmdline as GLIBC versions change".
Condition: Upgrading the SLX firmware from 20.1.2x to 20.2.x requires a 'fullinstall' firmware download in order to proceed.
Workaround: There is no workaround from EFA. The firmware download fullinstall must be carried out individually on each SLX device.
Parent Defect ID: EFA-8507 Issue ID: EFA-8507
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: Certain vlans are missed in configuration when stacks are created in quick succession within a script with no delay.
Condition: 10 stack creations without much delay leads to missing configuration. Trunk Subport update is not generated from neutron. Issue is seen with only one controller and not seen when more delay is introduced between stack creations. Trunk also remains in DOWN state.
Workaround: Workaround is to have delay between stack creation.
Recovery:

Remove the Trunk Parent port added to the VM and add it back again.

e.g.

Max-L2-ss3VirtIoVM2_Test1==> VM Name

Max-L2-ss3VirtIoTrunkPort2_Test1 ==> Parent Port of the Sub Port that is down

openstack server remove port Max-L2-ss3VirtIoVM2_Test1 Max-L2-ss3VirtIoTrunkPort2_Test1

openstack server add port Max-L2-ss3VirtIoVM2_Test1 Max-L2-ss3VirtIoTrunkPort2_Test1

Parent Defect ID: EFA-8512 Issue ID: EFA-8512
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: On SLX there can have partial config of neighbor under router bgp. The "show running command router bgp" from SLX shows invalid command "neighbor pg1" (Assume the bgp-group name is pg1). There's no corresponding command to delete this.
Condition: It's found if issue netconf RPC to SLX device with BGP peer group delete operation which the peer-group does not exist, SLX will create the invalid "neighbor pg1".
Workaround: Under some admin-down device scenario, avoid delete the same bgp-peer more than once.
Recovery:

On SLX use the following commands to get rid of the partial bgp-peer.

SLX(config)# router bgp

SLX(config-bgp-router)# neighbor pg1 peer-group

SLX(config-bgp-router)# no neighbor pg1 peer-group

Parent Defect ID: EFA-8526 Issue ID: EFA-8526
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: VRF Update "centralized-router-add" fail with error "[x, y] are MCT pair. Update the VRF with both devices together as centralized routers"
Condition:

1) In a CLOS fabric setup with MCT pair of border-leafs, create VRF with routing-type as centralized and select MCT pair of border-leafs as centralized routers.

2) Remove one of the MCT pair border-leaf from the fabric

3) Add same/different border-leaf to the fabric and run fabric configure command

4) Wait for sometime and run VRF Update "centralized-router-add" operation to add newly added border-leaf as centralized router

Workaround: Run VRF Update "centralized-router-add" operation and specify both nodes of MCT pair border-leafs as centralized routers.
Recovery: No recovery is required.
Parent Defect ID: EFA-8535 Issue ID: EFA-8535
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: On a single-node installation of TPVM, after ip-change, EFA is not operational.
Condition: After IP change of the host system, if 'efa-change-ip' script is run by a different user other than the installation user, in that case, EFA is not operational.
Workaround: Restart k3s service using the command 'sudo systemctl restart k3s'
Parent Defect ID: EFA-8573 Issue ID: EFA-8573
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: In few cases, networks in EPG will remain in cfg-in-sync state even if they are created with partial success topology (MCT pair with one admin-up device and one admin-down device).
Condition:

The issue is seen with the below steps

1) Configure a fabric

2) Create Tenant

3) Create multi-homed portchannel

4) Bring one of the devices of the MCT pair(having the PO created in step 3) admin-down to create a partial success topology

5) Create EPGs on the partial success topology

Recovery: Bring all the devices in admin-up state. It should push all the configs on devices and everything will be in cfg-in-sync.
Parent Defect ID: EFA-8950 Issue ID: EFA-8950
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.2
Symptom: SRIOV subports status is not coming to ACTIVE intermittently as the transition is not triggered from neutron service.
Condition: SRIOV subports status down with ss8/ss9 use cases
Workaround: In the SS8/SS9 usecases ,when we give little delay(2-3secs) for the network trunk set after sub port creation ,works fine. Sub ports coming to active state.
Parent Defect ID: EFA-8628 Issue ID: EFA-8628
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: Tenant does not contain any Ports, state of EPG is "vrf-delete-pending" and EPG Update "vrf-delete" fail with error- "EPG cannot be updated with tenant having no ports"
Condition:

Step1) Create Fabric and Tenant

Step2) Create VRF using Routing type as Centralized and Border-leaf devices as Centralized Routers

Step3) Create EPG using VRF created in Step2 with Physical Ports/Portchannel from leaf devices

Step4) Delete all the devices from Fabric and/or Inventory

Step5) Perform EPG Update "vrf-delete" operation

Workaround: No workaround is required.
Recovery: Delete and re-create EPG(s).
Parent Defect ID: EFA-8665 Issue ID: EFA-8665
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: VRF configuration present on Border Leaf devices
Condition:

1) Create Fabric and Tenant

2) Create VRF using Routing-type as "Centralized" and Border-leaf devices as Centralized-Routers

3) Create EPG(s) using VRF created in Step2 and Physical Ports/Portchannel from Leaf devices

4) Remove Leaf devices (having the Physical Ports/Portchannel used in the EPG) from Fabric/Inventory

Workaround: No workaround.
Recovery: Manually clean up VRF configuration from device
Parent Defect ID: EFA-8669 Issue ID: EFA-8669
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: EFA is not reachable after secondary node down/up of 8720. (kube-system pods are not responding)
Condition: This issue occurs only when the management cables are unplugged for under 10 seconds and plugged back in. What happens is that the EFA failover keepalived scripts, which does labeling and then switches the pod over to the other node, doesn't finish to completion causing the node labeling to not reflect right status where the VIP is assigned.
Workaround: Either keep the management ports unplugged for more than 20 seconds or reboot the tpvm.
Recovery: Reboot the tpvm.
Parent Defect ID: EFA-8701 Issue ID: EFA-8701
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: EFA becomes unresponsive and the OS reports that there is no disk space.
Condition: Galera is technology used by mysql to keep data in sync between nodes. Occasionally, such a transfer between nodes can fail, resulting in a log file of the failed transaction, and some binary data. These accumulate in /apps/efadata/mysql. EFA will tar.gz these logs hourly and delete; but if there is a rapid accumulation, this process can be slow, and under extreme circumstances, not complete before the next hour starts. Eventually, the system is unable to proceed.
Workaround: Delete files matching GRA_*.log in /apps/efadata/mysql. These are not needed by EFA.
Recovery: Delete files matching GRA_*.log in /apps/efadata/mysql. These are not needed by EFA. Kill all tar czvf processes working on GRA*.log files.
Parent Defect ID: EFA-8754 Issue ID: EFA-8754
Severity: S2 - High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: Shutting both mgmt and RME ports can occasionally cause EFA's kubernetes deployment to become stuck in an initializing state.
Condition: Turning off and on both mgmt and RME ports.
Recovery:

Restart kubernetes on both nodes of the cluster. As root user:

$ systemctl restart k3s

Parent Defect ID: EFA-8802 Issue ID: EFA-8802
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: Same epg will be allowed to update with different port properties value, however the first configured value will take effect
Condition:

1. Create L3 epg e1 with port-group P1 port-group property switch port mode trunk

2. Update epg e1 by port-group-add P2 and port-group property switch port mode as trunk

3. Update same epg e1 by port-group-add P2 and different port-group property switch port mode trunk-no-default-native

Above steps are applicable for all port properties like switch port mode , switchport-native-vlan-tagging, switchport-native-vlan, single-homed-bfd-session-type. No error is seen while updating with different port property

Workaround: Updating the epg with similar port property values since idempotency doesn't work
Parent Defect ID: EFA-8827 Issue ID: EFA-8827
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: CEP port will remain in "cfg-refreshed" state even after the DRC is successful.
Condition:

1. Create EPG with CEP ports.

2. Configure "MM enable on reboot" on the SLX and modify any of the CEP port properties.

3. Reboot the SLX.

4. MM triggered DRC will reconcile all the port properties but the CEP will remain in cfg-refreshed state.

Workaround: CEP in cfg-refreshed state can be removed from the EPG and readded to the EPG.
Parent Defect ID: EFA-8848 Issue ID: EFA-8848
Severity: S3 - Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.1
Symptom: When a configuration drift is introduced on a physical port on the SLX device, followed by a DRC on the device, the reconcile status displays "Portchannel" and "Vlan" reconciliation status as success even though the same were not drifted.
Condition:

Below are the steps to reproduce the issue:

1) Create Fabric/Tenant using MCT paired SLX devices.

2) Create an EPG using the physical port (P1) from an MCT SLX device (D1).

3) Introduce drift on port (P1) on the SLX device(D1).

4) Perform DRC of the SLX device (D1).

5) DRC executed successfully and all configuration pushed on the SLX device.

6) DRC output displays "Portchannel" and "Vlan" reconcilation status as success even though the same were not drifted.

Parent Defect ID: EFA-8966 Issue ID: EFA-8966
Severity: S4 - Low
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom: BGP peer group deletion fails when the deletion is attempted using a tenant not owning the BGP peer group.
Condition: Deletion of the BGP peer group which is not owned by an existing tenant.
Parent Defect ID: EFA-8967 Issue ID: EFA-8967
Severity: S4 – Low
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.0
Symptom:

Error during epg update ctag-range-add operation if different anycastIP is provided for new ctag range with same l2-vni already configured in EPG.

Ex.

Error : EPG: e1(11:11.11.11.1/24) shares network name with EPG: e1(10:10.10.10.1/24) should have same Anycast IP

Condition:

1. Create L3 EPG epg1 with ctag-1, l2vni-1, anycastIP1

2. Update EPG epg1 with ctag-2, l2vni-1, anycastIP2

For the same l2vni, anycastIP must be the same, the same condition will be verified as part of epg validation which provided above-mentioned error.

Workaround:

1. For same l2vni, anycastIP must be same.

2. Different l2-vni can be used for using different anycastIP.

Parent Defect ID: EFA-9014 Issue ID: EFA-9014
Severity: S2 – High
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.2
Symptom: 'efa fabric device add-bulk' returning Validate Fabric [Failed] due to "Missing Links"
Condition:

EFA as part of registering the device, enables the interfaces on SLX and has a wait time of 30 seconds before we check for LLDP data to infer links.

For some SLX platforms like 9740 the port bring up has a longer delay (Its been seen to be around 90 seconds in some corner cases).

Workaround:

There are 2 workarounds

1) register devices in inventory and have a delay of 1 minute before the next step which is to add devices to fabric. This will ensure EFA has the most current links.

2) Add devices to fabric without prior registration in EFA, follow that step by doing an `efa inventory device update --ip <fabric name>`

Recovery:

1) register devices in inventory and have a delay of 1 minute before the next step which is to add devices to fabric. This will ensure EFA has the most current links.

2) Add devices to fabric without prior registration in EFA, follow that step by doing an `efa inventory device update --ip <fabric name>`

Parent Defect ID: EFA-9045 Issue ID: EFA-9045
Severity: S3 – Medium
Product: Extreme Fabric Automation Reported in Release: EFA 2.4.2
Symptom:

1. "app-state" of some of the VRFs is shown as "cfg-refreshed" in "efa tenant vrf show" output.

2. The same VRFs whose "app-state" is shown as "cfg-refreshed" are shown as "unstable" VRFs in the "efa tenant epg show" output.

Condition:

1. Create an 18 node CLOS fabric.

2. Create multiple tenants (e.g. 14) tenants under the fabric created in step 1.

3. Create multiple VRFs (e.g. 400+) under the tenant created in step 2.

4. Create multiple EPGs using the VRFs created in step 3.

5. Check the "efa tenant vrf show" output to know the "app-state" of the VRFs.

Recovery:

1. Check the configuration drift per SLX device using the CLI "efa inventory drift-reconcile execute --ip <slx-device-ip>" to identify if the VRFs (which are in the cfg-refreshed state) are shown in the drift output.

2. If the VRFs are shown in the drift output for a given SLX device, then delete the VRFs from that particular SLX device.

3. Perform DRC (Drift and Reconcile) for the SLX device from which the unstable VRFs were deleted.

4. Steps 3 and 4 need to be done for all the SLX devices.