Our searches, both internal and external, can include versions of a product. However, there are often differences between proper product versions in our documents and how users actually search for those versions. For example, for this actual customer search problem, "sysview mq 9.2", we would want the following document returned: https://support.broadcom.com/download-center/solution-detail.html?aparNo=SO15546&os=z%2FOS However, it's not returned because it contains the text "9.2.0". What's interesting is that search does find other documents with the text "9.2.x". https://techdocs.broadcom.com/content/broadcom/techdocs/us/en/ca-mainframe-software/performance-and-storage/ca-sysview-performance-management/16-0/installing/prepare-for-installation/required-software.html https://techdocs.broadcom.com/content/broadcom/techdocs/us/en/ca-mainframe-software/performance-and-storage/ca-sysview-performance-management/16-0/installing/prepare-for-installation/supported-software-and-ca-products.html To summarize, a search for "sysview mq 9.2" returns 2 documents that contain the text "9.2.x", but not documents that contain "9.2.0", or even simply "9.2" (by dropping the trailing "0". We can improve our search tips article to encourage users to use wildcards when searching for versions, but is there a way to tokenize versions that include ".x" and ".0" so that it also works for just "9.2"? Note: This is just one example of many that would be affected by this.

Question: After a new SearchUnify product release, users are experiencing issues with SearchUnify. Links appear to be broken, Images are not rendering, Content Sources are missing and Search Bar is missing. Environment Production Reported product version Mamba'21 Resolved in version Mamba'21 Module Others Answer Users are advised to clear the browser cache and cookies after every product release to get rid of the old CSS and js saved on there local machine (cached). If issue persist, please access the web application in a different browser to rule out any browser issue. User can also do a hard refresh to see if content is rendering properly after the product release. Process of hard refresh is different for different browsers and OS: Chrome and Windows a) Hold down Ctrl and click the Reload button. b) Or Hold down Ctrl and press F5. Chrome and Mac a) Hold down Shift and click the Reload button. b) Or Hold down Command, Shift and click the ‘R’ key. For other browers and OS, you can refer to https://fabricdigital.co.nz/blog/how-to-hard-refresh-your-browser-and-clear-cache

Please find release notes for recent enhancements and updates in the community. 1. Case Form Pre-filled text in the Description field: As per Support team's suggestion, we have added pre-filled text in Description field of the Case Form. Idea is to capture as much details at the time of case creation so that back and forth between Support team and users is minimized. Content Source - Mention the content source. Write "NA" if this is not relevant to the case Search Client - Mention the search client, URL preferred. Write "NA" if this is not relevant to the case Product Version - Mention the current version, for example - Mamba 21, Colubridae 20 2. Digest Email We have updated subject of the Digest email to the format “Community Digest | Date Month, Year”. In case you're not subscribed to the Digest, please go to Settings section and enable it from there. 3. Community Functions Fixed notifications error on home page: Many of the users had reported that notification counter was not functioning properly. Notification count would not decrease, despite a user having read the notifications. Its been fixed in this release Added labels for Last Reply to posts: We have added a new label of "Last Reply" to the posts. This will help users to distinguish between the actual date when post is created and the last replies on the post.

Question What is the best way to exclude multiple pages from search when they are not in the same folder structure? Environment Production Reported product version M'21 Resolved in version M'21 Module Content Sources Answer The best way to exclude the docs while crawling would be to add all the URLs to the "Should not crawl" section. Below are the steps: Navigate to Content Sources in Admin Panel Edit website type Content Source from which you want to exclude the docs. Navigate to Rule tab and click "By Filter" Add pages in "should not crawl" section. This will prevent these pages from being indexed and shown in search result pages. Note: These URLs will also be excluded by auto crawlers that are running after a given frequency. Also, this feature is only available for website type content source.

Question: What is included in an Analytics subscription? Environment Production Reported product version Mamba'21 Resolved in version Mamba'21 Module Others Answer You can subscribe to Daily, Monthly, or Weekly Analytics report by following the steps mentioned in https://docs.searchunify.com/Content/Alerts/Analytics-Subscription.htm Below reports are segregated by each Search Client and included in the Analytics Subscription Top search queries - Click here for more info. Top search queries without result - Click here for more info. Top Conversions - Click here for more info Results with the least number of clicks Discussion ready to become help articles - Click here for more info.

Problem Statement: Why some of the documents are tagged as French in SearchUnify results when they are written and published in the English language? Users are not able to filter few articles based on language? When they select a language filter, corresponding articles are not shown in the search results? Why is the language filter for our doc site not coming? Environment Production Reported product version M'21 Resolved in version M'21 Module Search Client Causes: Content in the community is tagged to the French language. Articles are missing language metadata or not tagged to the correct language metadata. Language metadata information of the articles is not crawled Solution: Please check if the content present in your community is in the English language and not in the French language. SearchUnify crawlers use REST APIs to fetch content from different communities. Please click here to understand how to configure metadata for search results. If few articles are missing for a given language filter, please check whether articles are tagged and crawled with correct language metadata. For other causes of missing articles in the search result, please click here Make sure crawling is not stuck and all the content with correct language metadata is crawled.

Problem Statement: Issues in generating access and refresh token while setting up SearchUnify API Environment Production Reported product version C20 Resolved in version C20 Module Developer Guide Causes: Username/password used is incorrect Client ID and Client Secret is not Base64 encoded. Scope of the user is none Breach of API limits Solution: The first step to use SearchUnify APIs is to generate an access and refresh token. To generate tokens please refer to https://docs.searchunify.com/Content/Developer-Guides/curl.htm While running a command to generate or refresh access token below things should be considered: Use the same username and password that you used to log into the SU admin portal. Client ID and Client Secret have to be base64 encoded. Follow the mentioned link for encoding data: https://www.base64encode.org/ While encoding, the combination of the Client ID and Client Secret has to be separated by ":" (Client ID: Client secret). Please see the below screenshot. User scope should be 'Analytics' in the admin panel. For more information on how to change the scope please refer to mentioned doc https://docs.searchunify.com/Content/Manage-Users/Change-User-Scope.htm If the user gets an "API limit exceeded" exception, you can contact SearchUnify customer support for further analysis.