Release Notes for Mindbreeze InSpire
Version 2019 Fall Release
Mindbreeze GmbH, A-4020 Linz, 2020.
All rights reserved. All hardware and software names used are trade names and/or brands of the respective manufacturers.
These documents are strictly confidential. The transmission and presentation of these documents alone does not establish any rights to our software, our services and service results or any other protected rights. The passing on, publication or duplication is not permitted.
For reasons of easier readability, gender-specific differentiation, e.g. user, is not used. Corresponding terms apply to both sexes in the sense of equal treatment.
Innovations and new features
Personalized Relevance based on Machine Learning
This plugin can be used to influence the Mindbreeze relevance model (also called document boosting). This includes the search behavior of the users. The basic functionality of this plugin is that documents that have been clicked more often (open/preview action) will be displayed higher in future (identical) search queries.
Query Performance Tester
The Query Performance Tester can test the search performance and expected search hits automatically and reproducibly. The most important features are:
- Testing with Custom Search Apps
- Creation of test plans (what is searched for, what are the expected results)
- Parameterization of test runs (user, number of parallel searches, number of iterations)
- Monitoring of Test Runs
- Detailed test run statistics
Advanced Query Reporting
With the help of the Personalized Relevance Transformer Plugin reports about the search behavior can be created. Detailed search statistics and a timeline with filter functionality are provided.
Enhancements in the connectors
Microsoft SharePoint Online Connector
The Microsoft SharePoint Online Connector has received many general improvements. The most important ones are:
Site Discovery: Instead of configuring all pages individually, the crawler automatically recognizes the pages of the instance.
Delta crawling: The crawler now supports delta crawling. This means that after the first crawl run, only the changes since the last crawl run have to be retrieved. This significantly improves performance.
ADFS: The crawler user can now log on using ADFS.
Web Connector: Sitemap Download with Kerberos authentication
Kerberos Authentication optionally uses the Negotiate protocol to authenticate HTTP requests. The Web Connector is thus able to index Web pages that can only be accessed with Kerberos authentication.
The Notes Connector now supports Notes version 9. In addition, the Connector can retrieve DIIOP information via HTTPS.
No further configuration is required on the connector side.
Reindex without restarting the whole node
The "Reindex" action in the "Services" menu only restarts the affected index.
This does not affect the other services.
Improved status display during export, stability with large amounts of data
Export improvements make the status and export progress more visible.
Additionally, the stability when exporting large amounts of data has been improved
LDAP-Cache support for forest-trusts
LDAP cache now also supports ActiveDirectory domains linked with forest trusts.
Automatic fallback to producer/consumer nodes
If an index cannot be reached during a search in a producer/consumer setup, this new function automatically forwards the search query to other nodes. This improves the availability of the search, even if no load balancer is used.
- Security: Arbitrary host header accepted
- Java Security Update 18.104.22.168: Important: CVE-2019-2945,CVE-2019-2949,CVE-2019-2962,CVE-2019-2964,CVE-2019-2973,CVE-2019-2975,CVE-2019-2978,CVE-2019-2981,CVE-2019-2983,CVE-2019-2987,CVE-2019-2988,CVE-2019-2989,CVE-2019-2992,CVE-2019-2999
- Java Security Update 22.214.171.124: Moderate: CVE-2019-2745, CVE-2019-2762, CVE-2019-2769, CVE-2019-2816, CVE-2019-2842, CVE-2019-2786
- jQuery Update to 1.12.4
- app.telemetry Update to 18.3.60
- Apache Tika Update to 1.22
- Centos 7.7 Security Updates Host (CVE-2018-14647,CVE-2019-5010,CVE-2019-9740,CVE-2019-9947,CVE-2019-9948,CVE-2018-18074,CVE-2018-14348,CVE-2018-5741,CVE-2019-6470, CVE-2018-12641,CVE-2018-12697,CVE-2018-1000876,CVE-2019-3880,CVE-2016-10739,CVE-2018-15473,CVE-2018-16838,CVE-2019-3811,CVE-2018-16842,CVE-2018-1122,CVE-2018-0495,CVE-2018-12404,CVE-2018-20060,CVE-2019-11236,CVE-2018-20532,CVE-2018-20533,CVE-2018-20534,CVE-2017-14503,CVE-2018-1000877,CVE-2018-1000878,CVE-2019-1000019,CVE-2019-1000020,CVE-2018-0734,CVE-2019-1559,CVE-2018-15686,CVE-2018-16866,CVE-2018-16888)
- CentOS 7 Security libssh2: Moderate CVE-2019-3862 ,CVE-2019-3858,CVE-2019-3861
- CentOS 7 curl Security (CVE-2018-14618)
- CentOS 7 Kernel Security: CVE-2018-16871, CVE-2018-16884, CVE-2019-11811, CVE-2019-11085
- CentOS 7 Kernel Security: CVE-2018-20856, CVE-2019-3846, CVE-2019-9506, CVE-2019-10126
- CentOS 6 Kernel Security: CVE-2018-9568, CVE-2019-11810, CVE-2019-14835)
- CentOS 6 Security (Moderate): openssl (CVE-2019-1559)
- CentOS 6 Kernel Security (Important): CVE-2017-17805, CVE-2018-17972, CVE-2019-5489, CVE-2019-1125
- Centos 7.7 Security Updates Container (CVE-2018-14647,CVE-2019-5010,CVE-2019-9740,CVE-2019-9947,CVE-2019-9948,CVE-2018-18074,CVE-2018-14348,CVE-2018-5741,CVE-2019-6470, CVE-2018-12641,CVE-2018-12697,CVE-2018-1000876,CVE-2019-3880,CVE-2016-10739,CVE-2018-15473,CVE-2018-16838,CVE-2019-3811,CVE-2018-16842,CVE-2018-1122,CVE-2018-0495,CVE-2018-12404,CVE-2018-20060,CVE-2019-11236,CVE-2018-20532,CVE-2018-20533,CVE-2018-20534,CVE-2017-14503,CVE-2018-1000877,CVE-2018-1000878,CVE-2019-1000019,CVE-2019-1000020,CVE-2018-0734,CVE-2019-1559,CVE-2018-15686,CVE-2018-16866,CVE-2018-16888)
This version also includes the following changes:
- VocabularySynonymTransformer “loses“ original term when transforming alphanumeric queries
- jQuery is exported globally which can conflict with existing jQuery versions
- Query Performance Tester: support for Expectations and collecting statistics
- Query Performance Test Plugin with React-typescript UI (mit workspaces) is available
- Excel Export IllegalStateException: The maximum number of Cell Styles was exceeded. You can define up to 64000 style in a .xlsx Workbook
- Query Plugins: PersonalizedRelevanceTransformer does not work on position 2 (after ReplacementTransformer)
- Sample Texting Size of Export is not matching the Export Popup Sample Texting Size
- AccessConrolEntry action REQUIRE is supported
- Microsoft Sharepoint Online Get lists one-by-one if fetch of all lists fails
- StopcharclassTokenizer is not symmetric to RE2Tokenizer wrt. numeric/nonnumeric (3D is 3,D in numeric case)
- DocumentBoosting is 100% but not all DocumentBoost apply
- Term2Document Boosts fails completely when CSV contains entry with empty Key column
- Option for activating SampleTexting on export
- Atlassian Confluence Connector - unable to get character at index java.nio.charset.MalformedInputException
- Overwrite Documentum ACLs in configuration to allow repository wide role based access
- Documentum Crawler “recycle bin“ support
- Stemmer Transformer does not use boosting and has a bad effect on the hit quality.
- Facets show more" feature malfunction when the search result is loaded first time"
- mandatory_doc_info_schema_revision not patched on forceDocInfoRevision
- If a filter node value is identical to another node in the tree, these nodes have the same state
- mescontrol <url> repairreferences is available
- /find servlet is available even if the producer query service is disabled
- `Phrase` in Series is interpreted as a term of relevance
- GSA Feed Adapter: login form csv POST, cookie is not used anymore at http
- GSA Feed Adapter: Funktion Invalid document deletion schedule
- TermLexicon Merge Issue w Files > 4 GB
- Additional ReferenceTable for Property from IndexConfig, MesControl Mode to create for existing documents
- Index does not respond to "exit" command during "checkconsistency" commandWeb Connector - Accept-Headers and HTTP Request Header incorrectly processed when colon or semicolon is included
- Microsoft SharePoint Connector GetWebApplicationPolicy timeouts
- Make GSA Feed Adapter soTimeoutMs configurable
- Timeout with leading Asterisk-Pattern / MetadataTransformer
- GSA search transformer sorting on case
- Safari - MMC - Scrolling using the touch pad is not possible (only for iframes)
- Update Servlet needs too much memory when uploading Update
- LDAP cache is never checked for completeness after initial generation.
- Microsoft Exchange Connector support for Search Filter extensions
- Filesystem API with asynchronous operations for transferring files, directories
- GSA Feed Adapter - Following Patterns URLs are not getting processed when pushing feed twice
- Add a loading bar in the pdf preview window
- Indicate that facet counts are not total due to timeout
- Oauth Resource Proxy does not work in Microsoft SharePoint 2019 / Windows Server 2019 / IIS 10
- InSpire Config reloaded regularly with 3 nodes setup
- Microsoft File - SMB2 nt-authority\authenticated-users not part of Document ACL
- Recoverable inconsistency should not render index readonly after checkonsistency (Indexing is allowed)
- Configurable message while loading PDF Viewer
- VM args of Java sandboxes should be appended so that the generic vm args can be overwritten
- mesnode: verbesseres Logging
- Web Connector - converting IRI to URI with canonical links
- Table editing context menu does not work for MMC Search Experience / Suggested Sites
- Only the first DNS name / prefered LDAP server from the global LDAP server properties is effective
- No query optimization possible using category/fqcategroy for indexes with multiple categories
- Salesforce Connector uses obsolete api URL (www.salesforce.com instead of login.salesforce.com)
- Search Response can contain several facets with the same id
- Update is currently not recoverable if the cache directory is corrupted: Permanent Cache update failed ... Wrong index checksum, store was not closed properly ...
- Salesforce Connector: support for sandbox crawling (test.salesforce.com test system)
- Web Connector - Encode special characters in URI as %hex
- Optional resolution of email in Saleforce Authorizer and use of Query Plugin Keytab.
- Crawling sitemap OutOfMemoryError with Insite profile
- During constant syncdelta document is not inverted (missing from inverter log)
- Index Consistency Check should not set index to ReadOnly if the inconsistency is recoverable
- Connection Keep-Alive in RPC calls via HTTP
- Rebuild BTree process deletes documents with empty content from Document Table
- CookieManager Race Condition bei LoginHandler login
- InvertedIndexConsistencyChecker - Check that Inverted Index is Consistent w/ Storage
- WebCrawler Reset Session Before Login
- Talend 7.2 does not work with DataIntegration Connector
- Gunpowder support for grpc-web-text Protobuf API
- CachingPrincipalResolutionService: When several individuals match alias names, only one of them applies to the containers
- Crawling Public Folder Mailboxes
- Stemmer Vocabulary multicolumn CSV robustification (Vocabulary is not used to generate synonyms given same stems)
- Mindbreeze Management Cernter kann externe Ressourcen laden mit URL Pfad /ext/resources von /data/resources/mmc/wwww/.
- Download of Microsoft office documents that are rendererd to pdf are downloaded as application/pdf
- DocumentInfoAccessor for LookupService 37 gets docids for bucket 60
- Client Service: JSESSION Cookie cannot be deactivated currently
- Sync after deleteBuckets - not all transmitted buckets findable
- Outer CompositeDocumentInfoAccessor does not know if docid has been deleted or does not exist
- Sorting Results by custom date property uses monthly granularity (aggregated value representation)(1 result in wrong order - same order:num value)
- Pseudo Container Support for PushResolverComponent in the PrincipalResolutionCache
- Search reports can be generated by app.telemetry Client Service Query Log.
- Microsoft File Connector: Sortierung der Verzeichnisse/Dokumente nach Datum beim Crawlen (Neue zuerst crawlen).
- Filter C++ Sandbox2: errors are not propagated correctly to filter (always only std::exception in error text)
- DeleteBuckets deletes unused thumbnails or binary_contents
- percent_zone_boost and proximity improvement for Or query with phrases
- Index Sync (with large Microsoft SharePoint Index) hangs for hours with 100% CPU
- IBM Lotus Notes Connector should try database connection building with DbDirectory.
- Unfiltered items are transformed before skipped in InvertDocument::invert
- Google Drive Stoppt Crawl Run bei HTTP 403 reason:cannotExportFile
- Basic Login with umlaut in password does not work
- StringTable ERROR: Index consistency: unknown string id: 13" after rebuildbtrees"
- HL7v2 Connector throws parsing warning on delete message
- `unique(args...)` emits evaluated args only if not already emitted
- Microsoft File Connector / STATUS_INSUFFICIENT_RESOURCES
- Term lexicon based compound term split
- Google Drive Stoppt Crawl Run bei HTTP 500 message:null
- Inversion duration very long for ER rules with many occurrences
- Connector uses TLS1.1 which is no longer supported by salesforce.com
- QueryExprParserService is not instrumented correctly in app.telemetry
- WebRest mistakenly packages Spring 2.5 and Spring 3.0 at the same time
- Add rh-python-3.6
- PDF-Preview Title not working (displayed as "[Object]") if requested Format: [PROPERTY]
- In Suggest/Autocomplete it is possible to select a category