Brainspace Frequently Asked Questions

Ingest Questions:

Q: Why does the UI take a long time to change from the Saved Search screen when doing an Incremental Ingest?

A: When doing incremental builds. there is an issue where a paging process begins at the Brainspace API level that goes through and tries to determine the number of documents that need to be added through the selected save search. This process runs and is only visible in the brainspace.log. Once it is completed, the UI will be updated to reflect the ingest phase.

Our recommendation, for incrementals, is to use a saved search that includes only the documents to be added. It helps to reduce this issue on larger workspaces. However, since you have gotten this issue, the API is currently in the pre-ingest phase and it is best to allow it to complete. You can download and view the brainspace.log through the UI, the steps are outlined here: Basic Troubleshooting - Download and review the Brainspace.log. The brainspace.log will be the way to confirm that it is in the pre-ingest phase (you'll see a bunch of entries showing "Paging xxxxx out of yyyyyy").

We do currently have a request in with our dev team for improvements to the UI to show that it is paging instead of sitting with no updates on screen.

Q: Does Relativity^® 2-Factor Authentication work with the Relativity^® Connector? Does it work with the Relativity^® Plus Connector?

A: The classic Relativity^® connector uses direct connections to the Relativity^® SQL server and API and consists of Brainspace storing and using the authentication credentials on your behalf. Because of this, 2FA is not possible with the classic connector. The reason 2FA works with the Relativity^® Plus connector is because it utilizes the OAuth2 authentication with Relativity^®.

Build Questions:

Q: How do I make sure my build went well and the dataset is ready?

A: Check that the Dataset is showing as Active on the Datasets page.

Q: Why is my build showing Queued?

A: A build could be in a queued state due to another build currently running or if the Analytics Service is not running. To check if the service is running do the following command as root on the Analytics Server:

systemctl status brainspace-analysis

If you find the service is in a failed state then restart it with systemctl restart brainspace-analysis

If you still are having issues please contact [email protected] and include the brainspace.log and analysisserver.log.

Connector Questions:

Q: What is the default path in the DAT Connector?

A: By default you will use a . (period), which translates to /data/brainspace on the back end. If you would like to use a different directory it will need to be a directory in /data/brainspace and in the UI will be put in as .<your defined directory>/.

General Questions:

Q: What are checks that can be done for looking after system health?

A: The PDF found at the follow page discusses System Health Checks Brainspace System Health Checks.

Q: What is the minimum number of documents required to build a dataset?

A: On the initial build there must be a minimum of 300 documents with extractable text. Incremental builds can have as little as 1 document.

Q: Is there a list of common errors and solutions?

A: Yes! This list can be found at Brainspace Error Resolution Troubleshooting Table.

Q: Are Clawbacks supported?

A: No, clawbacks are not supported.

Q: What do messages like this mean: [ServerherdThread] WARN AsyncServer Failed to ping server: https://X.X.X.X:1604 [Connection refused (Connection refused)]

A: These messages mean that the Application is unable to communicate with the Analytics Service on either the Analytics or On-Demand. To resolve login to either of those servers and make sure the analytics service is running, if not restart it with “systemctl restart brainspace-analysis”

Q: Is there a minimum number of documents necessary to form a cluster wheel?

A: Brainspace needs a minimum of 300 documents to build a dataset. But these 300 documents must contain diverse enough text in order to build a brain, so in some cases, more documents may be required as our brain determination is based off of vocabulary.

Upgrade Questions:

Q: If I were to upgrade my Brainspace instance, why won't a supported OS version such as Red Hat 8/9 work?

A:  The Brainspace application (including JDK we leverage) are packaged as RPMs and have dependencies. That means they will not install unless those specific dependencies are satisfied, including the version.  More specifically, we build a custom Yum repository that is built for RHEL/CentOS 7 and is not recognized on a RHEL 8/9 system.

Example:

>> rpm -qpR brainspace-platform-6.5.4.x86_64.rpm 

dejavu-fonts-common 2.33-6.el7 

dejavu-lgc-sans-fonts 2.33-6.el7 

dejavu-lgc-sans-mono-fonts 2.33-6.el7 

dejavu-lgc-serif-fonts 2.33-6.el7 

dejavu-sans-fonts 2.33-6.el7 

dejavu-sans-mono-fonts 2.33-6.el7 

dejavu-serif-fonts 2.33-6.el7 

fontpackages-filesystem 1.44-8.el7 

fontconfig 2.13.0-4.3.el7 

postgresql96 9.6.11-1PGDG.rhel7 

postgresql96-libs 9.6.11-1PGDG.rhel7

In summary, if you were to upgrade to either a newer Red Hat version or different OS, your instance of Brainspace will no longer function properly.