Applies To:

Show Versions Show Versions

Supplemental Document: Release Information: F5 Application Connector 1.1.1

Original Publication Date: 12/20/2017

Application Connector Release Information

Version: 1.1.1
Build: 8.0

Known Issues in Application Connector v1.1.x

Functional Change Fixes

None


Application Connector Fixes

ID Number Severity Solution Article(s) Description
680645-1 1-Blocking   Upgrading BIG-IP to 13.1.0 causes Application Connector Service Center to stop working
683594 4-Minor   Service Center Log Verbosity Changes

 

Cumulative fix details for Application Connector v1.1.1 that are included in this release

683594 : Service Center Log Verbosity Changes

Component: Application Connector

Symptoms:
Setting the Service Center Log Verbosity to Debug or Silly in versions 1.1.0 and prior could cause the BIG-IP UI to become unresponsive under heavy Service Center API request load. The Service Center Log Verbosity in prior versions supported Error, Warn, Info, Verbose, Debug, Silly. The Debug and Silly levels have now been removed as user options in version 1.1.1 to reduce the amount of possible logging.

Conditions:
Under extremely heavy Service Center API activity, the BIG-IP UI could become unresponsive if the log level verbosity was set to Debug or Silly.

Impact:
Log levels for Service Center Version 1.1.1 are restricted to those noted above.

Workaround:
The maximum log level for users is now Verbose the default remains unchanged - Info.

Fix:
Service Center version 1.1.1 supports 4 log verbosity levels - Error, Warn, Info, Verbose


680645-1 : Upgrading BIG-IP to 13.1.0 causes Application Connector Service Center to stop working

Component: Application Connector

Symptoms:
After upgrading a BIG-IP configured with Application Connector to version 13.1.0, Application Connector Service Center status turns red and crashes repeatedly. /var/log/restjavad.0.log has the following error being logged repeatedly:

Aug 28 16:43:48 BIG-IP1 err sdmd[4252]: 018e0018:3: pid[14383] plugin[/Common/application_connector_plugin.application_connector_ext] throw er; // Unhandled 'error' event

Conditions:
This occurs when Application Connector Service Center is configured on a BIG-IP, and that BIG-IP is upgraded to version 13.1.0

Impact:
Application Connector will stop functioning.

Workaround:
There is no workaround. Avoid upgrading BIG-IP to version 13.1.0 when running Application Connector.



Known Issues in Application Connector v1.1.x


Application Connector Issues

ID Number Severity Solution Article(s) Description
672381 3-Major   Upgrade process will cause outage of Application Connector
653066 3-Major   Pool member status is Red on the standby, long failover times
651108 3-Major   Only Round Robin load balancing is supported with Application Connector nodes
650368 3-Major   No persistence records for traffic going through Application Connector
647157 3-Major   Application Connector can be assigned to a traffic group different from the virtual servers it serves
697418 4-Minor   Service Center 'Failed To Open' stat not accurate
682484 4-Minor   Rollback of Service Center from 1.1.0 to earlier release prevents service center from being deployed
656795 4-Minor   Source address persistence does not work
649410 4-Minor   You can only run one instance of the Application Connector Service Center
647183 4-Minor   Red iLX error banner is displayed while installing the Service Center RPM
646849 4-Minor   If the Application Connector Proxy loses connection to the BIG-IP, the GUI does not display the configuration

 

Known Issue details for Application Connector v1.1.x

697418 : Service Center 'Failed To Open' stat not accurate

Component: Application Connector

Symptoms:
The Service Center maintains a stat of the number of times a client connection failed to successfully open to a server. This stat is not accurate.

Conditions:
This will be encountered if the pool contains local servers.

Impact:
The impact is limited to the accuracy of the 'Failed To Open' statistic in the Service Center.


682484 : Rollback of Service Center from 1.1.0 to earlier release prevents service center from being deployed

Component: Application Connector

Symptoms:
If you try to roll back the Service Center from 1.1.0 to an earlier version by un-deploying the service center, uninstalling the RPM and then installing the older version, you will be unable to deploy an instance of the older version.

Conditions:
This occurs when attempting to roll back to an older version of the service center.

Impact:
Deployment of the service center will fail, status will be red.

Workaround:
If you wish to roll back to an earlier version, perform the following steps.

1. Disconnect all proxies from the service center
2. In iApps :: Applications LX, Un-deploy the service center
3. In iApps :: Applications LX, delete the template for the service center.
4. In iApps :: Package Management LX, uninstall the RPM
5. At the command line, run the following commands to clear rest storage completely:
IMPACT OF THIS STEP: this step will clear the iAppLX storage completely, if you had any other iAppLX applications installed they will be removed.

rm -f /var/config/rest/rpm/__db*
/opt/bin/rpm --rebuilddb --dbpath /var/config/rest/rpm
bigstart stop restjavad restnoded
rm -rf /var/config/rest/storage
rm -rf /var/config/rest/index
bigstart start restjavad restnoded

6. Install the new RPM and deploy normally.


672381 : Upgrade process will cause outage of Application Connector

Component: Application Connector

Symptoms:
You are only able to interact with the Application Connector Service Center on the active device and this includes performing upgrades. You must upgrade the service center on the active device. Furthermore, the upgrade process will upgrade the Service Center on the Standby device, therefore a brief traffic outage to the cloud nodes will occur while the upgrade completes.

Conditions:
This will occur when upgrading the Application Connector Service Center in an HA environment.

Impact:
After the RPM is installed the iAppLX will be restarted, and traffic is momentarily disrupted during the restart.


656795 : Source address persistence does not work

Component: Application Connector

Symptoms:
When you configure source address persistence for an Application Connector Virtual Server, persistence does not work.

Conditions:
This occurs with Application Connector on the Application Virtual Servers.

Impact:
Source address persistence does not work. The only persistence method that works in this version is Cookie Persistence, and it requires an iRule in order to work properly.


653066 : Pool member status is Red on the standby, long failover times

Component: Application Connector

Symptoms:
When Application Connector is deployed with health monitors, the pool members will appear Red on the standby. When the active traffic-group fails over, it can take several seconds (up to 20) for the pool members to become green.

Conditions:
This occurs when Application Connector Service Center is deployed in a HA environment. The long failover times can occur with or without health monitors.

Impact:
The Application Connector Service Center is not running on the standby traffic-group, so health checks will not be successful on the standby device until it has started and connected to the Proxies. This can cause the failover to take up to 20 seconds to complete.


651108 : Only Round Robin load balancing is supported with Application Connector nodes

Component: Application Connector

Symptoms:
When you add cloud nodes provided by Application Connector to a pool using any load balancing algorithm other than Round Robin, load balancing fails and all traffic is directed to the first pool member in the list.

This may also either prevent traffic from going to the cloud nodes , or force 100% of the traffic to the cloud nodes, if the cloud pool members co-exist with local pool members in the pool.

Conditions:
Cloud nodes provided by Application Connector exist in a pool, and the pool's load balancing method is not Round Robin.

Impact:
Since cloud pool members from Application Connector have no pool member statistics, all of the statistical load balancing methods will fail because the stats will read 0. For example if the load balancing method is Least Connections, the first pool member will always have 0 connections associated with it so all traffic will go to the first pool member.

Workaround:
Always use Round Robin load balancing with cloud nodes provided by Application Connector.


650368 : No persistence records for traffic going through Application Connector

Component: Application Connector

Symptoms:
When you run tmsh show ltm persistence persist-records, there are no records even though cookie persistence is enabled and working.

Conditions:
This occurs if persistence is configured for the virtual server, such as Cookie Persistence, and you want to see the persistence records.

Impact:
LTM will show no persistence records, and there is no way to see persistence records from LTM. All traffic is going through an iRule and bypassing the persistence tracking in LTM.


649410 : You can only run one instance of the Application Connector Service Center

Component: Application Connector

Symptoms:
When you try to launch a second instance of the Application Connector Service Center, the second instance is added but it does not work. The status shows as red and when you click into it, you see the configuration of the first service center.

Conditions:
This occurs when attempting to create more than one Application Connector Service Center using the iAppsLX menu in the GUI.

Impact:
Two service centers appear in the iAppsLX menu but they both are the same service center. All additional service centers appear as red.

Workaround:
Only configure a single Application Connector Service Center per BIG-IP device.


647183 : Red iLX error banner is displayed while installing the Service Center RPM

Component: Application Connector

Symptoms:
After clicking the install button while installing the Application Connector Service Center RPM, while waiting for the install to complete you see a red banner that indicates an iAppLX timeout error.

Conditions:
This can occur rarely while installing the Application Connector Service Center RPM. The error is due to the connection timing out after a certain number of retries but the installation is continuing.

Impact:
The error indicates that the installation has failed but it is still continuing. Wait a few moments and refresh the page and the install will complete.

Workaround:
If you see this error, in all likelihood the installation is just taking longer than what the install page expects, wait 1 minute for the install to complete and refresh the page and it should indicate that the install completed.


647157 : Application Connector can be assigned to a traffic group different from the virtual servers it serves

Component: Application Connector

Symptoms:
In a HA configuration, if Application Connector Service Center's virtual IP is assigned to a different traffic group than the virtual servers it is serving, an outage could occur if the virtuals are active on a different device

Conditions:
This can occur if Application Connector Service Center is configured on one traffic group, while the virtual servers it is tunneling exist on a traffic group. If the application virtusl servers are active on a device where the Service Center is in standby mode, the virtual servers will be unable to pass traffic.

Impact:
When this occurs, a traffic outage will occur.

Workaround:
When configuring Application Connector Service Center, make sure that the Proxy Virtual Server is in the same traffic group as all of the Application Virtual Servers.


646849 : If the Application Connector Proxy loses connection to the BIG-IP, the GUI does not display the configuration

Component: Application Connector

Symptoms:
If an Application Connector Proxy is connected and authorized with a BIG-IP, a record will show in the "BIG-IPs" table showing that it is connected; however, if the proxy loses connection to the BIG-IP for any reason (network issue, dns issue, BIG-IP restart or reboot), the GUI will stop displaying the BIG-IP.

Conditions:
This will occur if the proxy loses connection to the BIG-IP for any reason.

Impact:
If you look at the proxy GUI it will appear that there are no BIG-IPs configured, nor any nodes, and that you need to configure everything from scratch.

Workaround:
If the BIG-IP was simply rebooting, you may be able to re-add the BIG-IP connection to the proxy, and BIG-IP will automatically re-authorize the proxy. Once this occurs, any pre-existing nodes that were configured on the proxy when the BIG-IP is reconnected.




*********************** NOTICE ***********************

For additional support resources and technical documentation, see:
******************************************************
Generated: Thu Dec 14 10:10:30 2017 PST
Copyright F5 Networks (2017) - All Rights Reserved

Was this resource helpful in solving your issue?




NOTE: Please do not provide personal information.

Additional Comments (optional)