In this Page
Live on May 2, 2015
Before You Begin
You need to clear your browser cache before you log into the latest SnapLogic Elastic Integration Platform.
See the appropriate documentation for your browser:
You will see an Upgrade Notice dialog informing you that you will be logged out immediately to complete the update process. Because the SnapLogic Elastic Integration Platform consists of multiple applications, you will likely see this message multiple times.
- Lifecycle Management
- Project promotions are now atomic. If any item fails during a promotion, the project is not promoted.Snaps
- Splunk Saved Search
- Splunk Writer
For information on Snap updates, see Snap Updates, 2015.
Any FeedMasters which are running will be automatically updated to the new version. If a FeedMaster is not running, the next time it is started, it will fail to initialize itself with the SnapLogic cloud and the auto update will fail. The workaround for this is to comment out the "server_type = feed_master" entry in global.properties and restart the JCC. This will enable the node to update itself. The server_type can then be changed back to feed_master and the JCC restarted. This workaround is required for the current release only; it would not be required for subsequent releases.
Ultra tasks will be automatically disabled if the pipeline fails to start or fails to run for two minutes after N attempts. This is to prevent faulty pipelines from overworking the system. The Max Failures property of the Ultra Task can be used to control the number of attempts before being auto-
disabled or the property can be set to zero if the task should never be auto-disabled.
For locally triggered tasks (those using the On-Premises Task URL), the HTTPS port (8081) on the JCCs is used to do the data transfer between nodes. This means that the Groundplex nodes no longer need to be configured so that they listen on 0.0.0.0 as described in the 'Snaplex Network Binding' section at http://doc.Snaplogic.com/configuration-options
(By default, the JCCs listen only on localhost for their HTTP port, but they listen on all network interfaces for the HTTPS port.)
Users should make sure the nodes can talk to each other on this port.
The status of completed child pipelines is not displayed in the running pipelines section when the parent pipeline has subsequent child pipelines.
In Dashboard, the Executing Pipelines section will only show those pipelines currently executing, and it if has children, which children are currently executing. If a parent pipeline has children which have completed, they will only appear in the Pipeline Execution History when the parent has completed, along with the parent if Group Pipelines is selected. If Group Pipelines is not selected, that list will show all completed pipelines, including any completed children of still executing parents.
MapReduce job fails when using a large JSON file for input data
Union Snaps require same field names in the same order
Potential issue with joining multiple or large input files
Pipeline generates an empty output file if the input file has fields with dots and brackets
MapReduce job fails without proper error if there is an empty value in input data
Promotion of a hundreds of assets can take up to 15 minutes to complete.
A promoted project will allow any members of a phase to read and execute the assets. As a work around, the org permissions can be removed, but then read and execute permissions for a group must be added to ensure a user can properly access the assets.