'Search' not working after upgrade to 8.3
We have upgraded from 7.2 to 8.3
As of now our ‘Search’ button in the top right to search for PEGA rules does not seem to be working.
Any ideas….
Accepted Solution
Pegasystems Inc.
US
You can either remove the jvm arg so that Pega defaults to those 4 node types which includes stream, search.
OR Include the additional node types as comma separated values to include search.
Eg: - DNodeType=Stream,Search,BackgroundProcessing,Webuser
Pegasystems Inc.
US
Hi Henry,
Check if your System-Setting->Search landing page shows Healthy AVAILABLE indexes with a valid Index node status as AVAILABLE.
Check if the below INFO messages were printed on JVM startup which confirms Full-Text is initialized with out any issues.
INFO - Initialized embedded full text search functionality for this node. Full-text search enabled for node XXXXXXXXXXXXXXXXXXXXXXX
NFO - Finished configuration of node for full-text search, with no errors.
Also, it's worth checking the things mentioned in the below article to start with.
Let me know if that helps to check further things.
Pegasystems Inc.
US
This shows that none of the rule/data/work indexes were available or reachable from the upgraded node.
Do you see an Index node status with index directory with status as AVAILABLE?
If NOT, you need to add one node as INDEX with a specific directory and submit and initiate Re-Index.
You don't have to ReIndex if there is an index directory already existing with proper search indexes, you could just map it and restart the node to see if it's full-text search is initialized properly. (look at the startup logs to confirm).
Also, make sure the other prerequisites like Stream node type & pyFTSIncrementalIndexer is running fine.
Veteran Careers First
US
Search Indexing
- So we added the Host Node ID and File Directory.
- The Node Status says online after server restart.
- We clicked on ‘Re-Index’ but nothing happened.
Stream Service
- The article you mentioned says to ‘add a stream service’.
- But we already have one listed with a status of NORMAL.
pyFTSIncrementalIndexer
- The article then also says to start the Queue Processor: pyFTSIncrementalIndexer.
- Currently ours says ‘Not Started’.
- When clicked it says: ‘cannot be started as there are no nodes of Search Type’
- The 3 dots to the very right has an option to ‘View Data Flow’
Data Flow Services Landing Page
- Here there is a drop down for ‘Search’ to add Data flow Nodes
- Nothing is listed here at the moment
- But there also is not an option to ‘Add’
Pegasystems Inc.
US
Pick the index node which was added to the search landing page and assign a node type "search" to it using below JVM arg.
"-DNodeType=Search"
NodeTypes classification is needed on 8.x versions- https://community.pega.com/sites/default/files/help_v82/procomhelpmain.htm#engine/node-classification/eng-node-types-ref.htm
By default, if you don't assign any node type, Pega will default the nodetypes to the below list. Hence, ensure you have the search nodetype also set to the index node added.
Startup Log message - INFO - Node is started without NodeType. Hence NodeType will be the default list: [WebUser, BackgroundProcessing, Search, Stream]
Also, the File Directory added as Index directory has indices node data? If yes, then you don't need to reindex. The rules/work/data will show available with certain index sizes in MB
Let me know how it goes.
Pegasystems Inc.
US
You would see a node in DataFlow Search type if there is a DataFlow.Search service started.
This service will start on JVM startup with if -DnodeType=search is assigned.
Pegasystems Inc.
US
It is a JVM argument, if you don't assign any node type, Pega will assign search nodetype along with WebUser, BackgroundProcessing, Stream
Can you share the JVM's startup logs here to confirm if search nodetype is assigned & search is initialized?
I also encourage you to go through the below link for all info about search & prerequisites on 8.x
Accepted Solution
Pegasystems Inc.
US
You can either remove the jvm arg so that Pega defaults to those 4 node types which includes stream, search.
OR Include the additional node types as comma separated values to include search.
Eg: - DNodeType=Stream,Search,BackgroundProcessing,Webuser
Veteran Careers First
US
It seems to be working now. Thanks for your help.
Search Indexing
- Added the Host Node ID and File Directory.
- The Node Status says online after server restart.
Stream Service
- Has a Node listed…
- Currently with a status of NORMAL.
pyFTSIncrementalIndexer
- Currently ours now says ‘Running’
Data Flow Services Landing Page
- Now has a Node listed…
- Currently with a status of NORMAL.
Decision Consultant Solutions
GB
From Pega 8.3 onwards, the "Add node" button will not be available and you need to classify the nodes using -DNodeType JVM argument only.
Have look on the below articles which might help you.
From Pega 8.3 onwards, the "Add node" button will not be available and you need to classify the nodes using -DNodeType JVM argument only.
Have look on the below articles which might help you.
Pegasystems Inc.
IN
Hi Henry,
Please watch below video tutorial covering all the concepts of search.
It shows value of JVM argument to add.
Thanks,
Abhinav