Release Notes for 12/07/2021

Pika Release Notes 2021.04.0

New Functionality

Solr Upgrade (Minute 00:42)
- Reworked the index backup and index restore to use Solr’s native backup/restore processes
- Compress backup files as a background process so that other nightly full update processes can proceed
- When the Solr searcher core dies while the indexing core is doing the full re-index, restarting the solr searcher would cause search results to empty out. The restart script was reworked to issue commands to solr after the restart to prevent the results from emptying out (patch deployed)
- Built a solr restart script for when the solr searcher core becomes unresponsive. Includes the above features if this happens during the full re-index
- Removed obsolete language boosting from index and code as well as some other obsolete or unused index fields
Advanced Search Form (Minute 4:01)
- The Advanced Search Form display query is now shown outside of the basic search input box to avoid implying basic searches of the advanced search form queries are equivalent.

Advanced Search Form urls will contain only populated parameters instead of many unpopulated (and thus unneeded) url parameters.
Searches with multiple options for the format facet chosen will keep each of the options displayed instead of just one selected option (Minute 7:03)

NoveList Information page provides the data retrieved from an ISBN (Minute 24:30)

NoveList series displayed in full record and search view now displays separately (Minute 19:53)

Grouped Work view page, staff section now displays the NoveList Primary ISBN associated with that work.
Series Page sharing options (Minute 20:47)
- Email
- Export to Excel
- Export to List

Background cron process to remove individual marc files from server when the record is no longer part of a collection to help conserve disk space
Overdrive download actions changed to use their new handling method (Minute 22:25)

Solr Upgrade related issues (Minute 8:19)
- Restore the default search term operator as AND instead of OR. eg The search phrase “my little pony” should result in searches of “my AND little AND pony” instead of “my OR little OR pony” (patch deployed)
- Grouped Work Id searching (via Keyword search type) will have the id match as the top result (parts of the id can coincidently match parts of titles)
Links on author names with diacritical characters returned no search results due to bad handling of the diacritical characters. Now returns expected search results. (patch deployed) (Minute 10:15)
Links on Uniform Titles with diacritical characters returned no search results due to bad handling of the diacritical characters. Now returns expected search results. (patch deployed)(Minute 10:56)
Links on subject headings where the subject phrase contained very common words (eg, for, with) would return no results. Now returns results for every work that has that subject heading.(Minute 11:21)
Overdrive author name now indexed in “last, first name” style consistent with MARC (Minute 13:45)

Advanced search for invalid or erroneous record number phrases using the Record Number search type will now give no results instead of solr error page.
NY Times list would not update when their API listed an ISBN as “none”
Users that can make browse categories can only add to their own library’s or location’s browse categories when using another library’s site.
The display of MARC in the staff view section of record views sometimes would not wrap so that the information was visible.
The tables in the Grouped Work view staff view section are often larger than the visible area. (record details, item details, scope details) Additional wrapping was added to improve the display for these tables
When 'no preference' or ‘none’ is the hold notification type for the user in Sierra, use the generic hold alert text (Minute 29:10)

Sierra: PIN length min and max length (Minute 17:05)
- Checks for self registration
- Checks when updating my account
Sierra: Allow email in MyAccount to be updated to null (Minute 17:54)

Novelist issues (Minute 25:32)
- Incorrect primary ISBN being reported by Novelist API
- Leading to incorrect book covers, incorrect series, and other discovery issues.
When Marmot’s primary DNS is down, calls to OverDrive API had connection timeouts which caused user actions with OverDrive titles to fail. The connection time out amount was increased to 5 seconds. This will have the side effect of pages loading slower when a user is logged in and the OverDrive API is unresponsive or the Marmot primary DNS is down. (Minute 15:29)

Sacramento Public Library - Change labelling for Materials Request for patron-facing pages to Recommended Titles (Minute 29:57)
Marmot enabled DKIM on Wake’s Pika server. DKIM is a way to sign the email coming from the server with a verified encryption key. This assures the recipient that it has originated from the intended domain and is not SPAM.
Updated theme for LION consortium (Minute 30:25)
New theme for LION library Wheeler
Anne Arundel library updated theme/branding (Minute 30:48)
Flatirons migration to barcode/PIN login (Minute 18:48)

Meeting Date:

Tuesday, 2021, December 7

Documentation Type:

Release Notes

Link to File:

Committees:

Discovery Committee