The inner-workings of Knowledge Management can be complex for an end-user or administrator to understand when new to the system following a transformation or upgrade. The following guide should assist with the basic and complex questions posed by consultants and customers. The guide is largely focused around Mid Tier but will evolve to include Smart IT
I have spent the past 10+ years working with, migrating and implementing Knowledge Management for BMC customers such as Lundbeck, Volvo, Serco, Aker and IKEA and wish to share the knowledge gained through such implementations.
I hope that this guide is useful, please feel free to provide feedback and/or any updates that you feel are required. If you do share or provide this information. If you feel that a section should be added to address a specific area of confusion, please make a request using the comments section at the base of the article.
IF SHARING THIS GUIDE - please ensure that you include a link to this specific URL (https://communities.bmc.com/docs/DOC-32647) so that the most recent updates are available to the end-user as this document will be continually updated. Please also do not claim this content to be your own, this guide took some time to author, I would appreciate a credit if this content is useful to be utilized elsewhere, and also to be asked before publishing.
The views contained and suggested within this document are my own and are not on behalf of BMC Software official guidance.
Table of Contents
Basic vs Advanced Knowledge Search
Incident Knowledge Search vs Standard Knowledge Search
Knowledge Article Content Indexing
Knowledge Content Weighting
Knowledge Search Results Display
Knowledge Search Results Selection
Knowledge Search Results Order
Knowledge RTF Editor Image Embedding
Knowledge RTF Editor Copying/Pasting Images
Knowledge RTF Editor Issues
Knowledge Attachment Indexing
|14||Knowledge Stop/Ignore Words|
|15||FTS Fortification with Remedy 9|
|16||Knowledge & FTS Common Issues/Problems|
|17||Knowledge Tab/Window Issues|
|18||Knowledge Management with Smart IT|
|19||Knowledge Article 'Last Modified Date/By' Fields|
|20||Knowledge Update Requests|
|21||Knowledge Status Transitions|
|22||Knowledge Process Recommendations|
|Date||Description of Changes|
|04/03/2015||Added section 13 concerning supported FTS Attachment Types for indexing.|
|11/05/2015||Enhanced section 4 to provide additional detail on content weighting/relevancy - also implemented some formatting enhancements for TOC and Version History.|
|13/05/2015||Added section concerning RKM 'Stop/Ignore Words' and consideration on what should/should not be included.|
|23/07/2015||Enhanced section 4 concerning RKM 'Weighting'.|
|31/07/2015||Added a new section on FTS Fortification.|
|10/11/2015||Added new section (16) for common FTS/Knowledge problems.|
|11/12/2015||Added new section (17) to describe inconsistencies in the way that Knowledge Articles open.|
|20/01/2016||Added new section (18) to discuss Smart IT with Knowledge Management.|
|29/02/2016||Added new section (19) to discuss the quirks around the 'Last Modified' fields in an RKM article.|
|21/04/2016||Added a new question/answer to the Knowledge 'Common Issues' section concerning the assignment field.|
|02/03/2017||Added new section (20) to discuss an illogical workflow for 'Update Requests'|
|25/05/2017||Added additional common problems|
|23/10/2019||Added new section (22) for process recommendations|
1 Basic vs Advanced Knowledge Search
“There are two search methods available Basic and Advanced. What are the differences between these? How do they work and what are the defaults.”
The Simple Search page enables you to perform simple searches. When you enter multiple query terms as search criteria, BMC Knowledge Management uses the Boolean ‘OR’ operator to query the knowledge base.
For example, if you enter “hardware software failure”, BMC Knowledge Management returns results that contain any of these terms because an ‘OR’ operator is used by the system by default, however the user may insert an ‘AND’ operator in-between the keywords to force a type of search which will only return results where all of the words are found, e.g. “hardware AND software AND failure”.
A Simple Search will by default include all Knowledge Sources that are accessible by the user (e.g. Change Requests will only be shown if the user has a Change Permission), however the default selected Knowledge Sources change if a Knowledge Search is invoked from an ITSM module, i.e. Incident. A Knowledge Config user is able to 'Disable’ any Knowledge Sources which an organisation do not wish to be searchable via Knowledge Search.
The Advanced Search page provides fields that you can use to define additional search criteria and narrow your search results, using any of the following Knowledge Article attributes:
- Knowledge Source (ability to exclude certain types of Knowledge Article).
- Knowledge Article ID, Language and Status.
- Knowledge Article Created Dates.
- Organisation (Company, Region, Site Group, Site).
- Operational Categorisations
- Product Categorisation
- Business Service
- Visibility Groups
After you set advanced search parameters, the Use Advanced Search Criteria check box appears in the Simple Search page. This enables you to change search words and include advanced search parameters without having to go back to the Advanced Search page.
Generally, from a process perspective it is recommended that users initially try a Simple Search, if too many results are found then the user can switch to an Advanced Search and define specific criteria.
For more information, refer to the following links (both require docs.bmc.com access):
2 Incident Knowledge Search vs Standard Knowledge Search
“The search methods can be evoked either from the incident module or knowledge module. What are the differences between these?” and “When a user evokes the search functionality from the incident management module how does that work.”
When a Knowledge Search is invoked from the Incident module (using the ‘Search Knowledge’ link under the left-hand Incident menu), the search interface will open as a floating dialog on top of the Incident Console.
The dialog is visually in ‘Basic Mode’ and the ‘Summary’ from the Incident is automatically placed into the keyword box – this alone demonstrates how important it is that Service Desk users consider carefully the value they use for an Incident Summary.
The ‘Use Advanced Search Criteria’ checkbox is automatically checked for the user and advanced criteria is populated with values from the Incident and therefore automatically filtering the Knowledge Results, as per the following example:
The Incident has the following categorisations:
When the user clicks ‘Search Knowledge’ from the Incident:
The ‘Advanced Criteria’ within the Knowledge Search is automatically populated with the categorisations (Product and Operational) from the Incident:
In addition, only Knowledge Article ‘Sources’ are checked (Knowledge Article Template types):
The automated population of these elements can filter the knowledgebase too thoroughly which can cause no knowledge results at all, in this context the user should uncheck the ‘Use Advanced Search Criteria’ checkbox to commence a ‘Basic Search’ based on the keywords only.
Once the user selects a Knowledge Article to view, a ‘Use’ button is available at the bottom of the form, clicking this will automatically add a ‘Relationship’ record to the Incident, associating it with the Knowledge Article. Once the user clicks ‘Close’, the ‘View’ counter is also incremented which can be reported on via an OOTB report within the ‘AR System Report Console’.
In addition, the Resolution Categorisations from the Knowledge Article are automatically added to the Incident ‘Resolution Categorisation’ dropdowns.
3 Knowledge Article Content Indexing
“What content is indexed in the knowledge article and what is the impact”
The following fields within the Knowledge Article are indexed:
- Article ID
- Product Cat Product Name
- Product Cat Model/Version
- All RTF Content Areas
When a ‘Keyword’ is entered, only the above listed fields are searched for inclusion of the keyword string. Some elements under ‘Additional Categorisations’ are only searched in comparison with any ‘Advanced Search Criteria’ that has been entered or automatically populated when accessing the Knowledge Search from another ITSM module.
It is worth mentioning that the ‘Keywords’ field can contain multiple words which require no delimiter, for example the following would be acceptable:
- “network laptop server protocol tcp”
Attachments to an article such as Word, Excel and other Office documents are also crawled by the FTS engine and are searchable.
4 Knowledge Content Weighting
“Weightage rules and its impact on search”
The OOTB configuration results in all of the above being of equal weighting of all indexed fields, however via configuration it is possible to alter the ‘Title’ and ‘Keyword’ weighting via the ‘Server Information’ Form:
Increasing the 'Title' value specifically to at least "4.000000" and the 'Keywords' field to "2.000000" tends to work efficiently in terms of raising the appearance of specifically matching articles within the search results, for articles that match elements of ‘Title' or 'Keywords', this avoids matches within the content having a higher relevancy rating than the summary of the Knowledge Article which is quite often the most relevant string/content.
I have published a BMC 'Idea' (RFE) to amend the default value of this field in an OOTB installation as this setting change improves the end-user experience in providing more relevant results -Change FTS Default for 'Title' and 'Keywords' Weighting
NOTE: After altering the above weighting values, it is recommended to reindex the Knowledge Sources to see the effect, this can be achieved using the 'Manage Knowledge Sources' option from the Knowledge Management Console, reindex each RKM source individually - this avoids having to run a full reindex via the 'Server Information' screen.
In addition to field weighting, other attributes of a Knowledge Article can alter its position within search results.
- Article View
- Article Use
Whenever a Knowledge Article is used (achieved by clicking the ‘Use’ button at the base of an article), a ‘Use’ counter is incremented. The same applies to whenever an article is opened and then closed using the ‘Close’ button, a ‘View’ counter is incremented.
An OOTB boost is given to articles based on their ‘View’ and ‘Use’ counters, the value of this boost if configurable via the ‘Application Administration -> Knowledge Management’ settings:
The OOTB default values for these fields has little effect on results, increasing these values to the following can give more of a notable difference to search results:
- View Boost Increment: 0.001500
- Use Boost Increment: 0.050000
It should also be noted that the 'Keywords' field shouldn't be populated with excessive keywords, it should only contain relevant synonyms for the core elements in the article title.
5 Knowledge Search Results Display
“Search results display - What content is displayed for each article, how it is display and what sequence or priority.”
The search results will also display the Article ID followed by the Article Title. The snippet of text that follows is configurable as explained below.
The icon to the left of the Article ID indicates the type of content, whether a Knowledge Article template or a record from elsewhere within ITSM (i.e. Incident, Tasks, Change Request).
One OOTB configuration option exists for the display of search results, these settings are configurable per Company and can be changed with ease via the ‘Application Administration -> Knowledge Management’ setting:
These options control the display of matching values within the Search Results. The following examples illustrate how each performs in a live system:
NONE – This setting excludes any element of snippet from the matching article:
EXCERPTS – This setting will not illustrate the sentence/paragraph that has matched the search keywords specifically, instead it will show an excerpt from the article or the attachment, whichever matches the search terms.
WORDS AROUND HITS – This setting shows the end-user the specific elements of the article that matched the search keywords. It concatenates (using * as the delimiter) any of the indexed fields that have matching content to give the end-user an idea of the context of the matching result. This is the most useful setting that we recommend for customers.
The way in which the above ‘Words around hit’ snippet is determined by the Lucene configuration, this is the underlying third-party ‘search engine’ that is used by AR System.
The sequence or ordering of the search results is determined by a Relevancy Ranking Algorithm of the Lucene search engine. It is an involved algorithm but in simple terms we can say that the Relevancy Score is calculated by considering the following:
- The Frequency of search terms in document - for example, if the word "oracle" is included in 8000 of 10000 documents, but the word "MySQL" is in a much smaller number of documents, the term "MySQL" is considered more relevant to the query than the term "oracle"
- The number of times a search term appears in a document - for example, if a document contains the word "oracle" five times in its description, that document is considered more relevant than another documents that might only contain the term "oracle" once.
- The length of the document text - if a search term is found in a document that contains a short amount of text, that document is considered more relevant than a document with a larger amount of text.
- Artificial "boost" factors – discussed under Section 4 under “Knowledge Content Weighting”
6 Knowledge Search Results Selection
“Search results display – selection of Knowledge Articles”
An explanation of searching indexed fields has already been given in section 3, however here is some additional guidance on how best to search Knowledge Management to gain the most efficient results display.
The OOTB ‘Basic Search’ will use an ‘OR’ operator, unless the user purposely adds an ‘AND’ operator between each keyword.
The following Booleans (with example results) are available during both Basic and Advanced searches, as follows:
In addition, the following characters can be used to refine the search:
7 Knowledge Search Results Order
“Search results display – order of display of knowledge articles”
An explanation of searching indexed fields has already been given in section 4, however this section will provide additional information.
As detailed, all fields have an equal ‘Weighting’ value of 1.0 in the OOTB configuration, if two articles match the keywords but one occurrence is in a boosted field, this article will appear before the other in the results. In summary, if the Knowledge ‘Title’ and ‘Keywords’ fields are deemed most valuable, their relevance value should be boosted above the OOTB value via configuration.
As per the docs.bmc.com resource,
- “Search criteria — According to the number of times that the search criteria was found in an article, the number of words in a searched phrase, and the order of words in a searched phrase.”
- Article boost parameter — Increases relevance when an article was found useful or viewed by users, as follows:
- o Use — You indicate that an article is useful to you by clicking the Use button that is available for that article.
- o View — You open an article for viewing from the search result list.
A combination of the above three elements determines the search results ordering of an article.
8 Knowledge/FTS Reindexing
“How and Why should I reindex Knowledge Management/FTS?”
Customers and consultants regularly find issues where a Knowledge Article cannot be found within a Knowledge/Global Search, or multiple versions are found when searching, some which are out of date, in some cases no articles are found at all.
Under certain circumstance, the FTS engine can cease its indexing function, generally due to infrastructure limitations or misconfiguration such as:
- Insufficient disk space on the server holding the 'Collection' folder.
- Insufficient memory (RAM) or CPU capacity on the FTS indexing server.
- Misconfiguration or Disabled Knowledge Sources (within Knowledge Management Console).
- Misconfiguration of wrongly allocated Visibility Groups.
- Misconfiguration of user permissions/Support Groups.
Once every element of configuration has been verified but FTS is still incorrectly indexing, the following steps will entirely flush your FTS index and refresh the Collection folder files, this can speed up your search requests as well as reducing the size of your Collection folder store.
The steps must be followed precisely and in the exact order as listed below. For servers with multiple FTS indexing servers (and multiple Collection folders), those servers should also be taken through this process.
- Stop your ARS services
- Empty the contents of the FT_PENDING Database table
- Empty your FTS /collection/ folder (remove all files within it, but leave the folder)
- Start your ARS services and wait for them to restart.
- Go to 'Server Information -> FTS Tab' and select 'Reindex'
- Wait for reindex to complete, then retest FTS.
- If still having difficulties, go to the 'Manage Knowledge Sources' screen, click 'Refresh' top right of it and verify that all state 'Ready'.
You can check on the progress of the reindex by checking on the records within the DB table 'FT_PENDING'. Query this table, eventually the table will be empty, of the bulk of records, indicating the reindex is complete.
9 Knowledge Reporting on Feedback
“How is the OOTB 'Knowledge Feedback' report calculated across multiple article versions?”
This OOTB report is calculated based on the average score across all versions of a Knowledge Article. You can test this by creating an article, submitting a Feedback score of 1. Then create a new version, submit a feedback score of 5. Then run your report and you'll find that it averages at 3.
10 Knowledge RTF Image Embedding
“Is there a limit imposed on the number of images that I can embed into my Knowledge Articles?”
Yes, there is a limit imposed on the number of images that can be embedded within each content area of a Knowledge Template, however it is important to note that these 'attachment pools' are separate to the main article attachment pool available on the 'Details' tab of each article (which has a limit of 25 attachments).
This section is applicable only to the images that you are adding directly to an article using the RTF editor.
Each RKM Template has differing limitations for the number of images that can be embedded, this number is determined by the number of attachment fields and pools within the core template Form - to inspect or customise these (to add more attachment fields), you must alter the following form from within Developer Studio.
Each of the above templates has varying content areas, for example the 'How To' template has 'Question', 'Answer' and 'Technical Notes', whilst the 'Reference' template has 'Reference' alone. Each content area has a different limitation on how many images can be placed into it.
- Known Error
- Error: 30
- Root Cause: 15
- Workaround/Fix: 15
- Technical Notes: 15
- Reference: 30
- Reference: 30
- How To:
- Question: 30
- Answer: 15
- Technical Notes: 15
- Problem Solution:
- Problem: 30
- Solution: 15
- Technical Notes: 15
It would not be recommended to increase the number of attachments (via customisation) as this has various implications:
- Will increase the size of the DB over time.
- Will slow the speed of display for articles as the images also must load.
- Will cause the article length to increase substantially, therefore requiring the user to scroll.
If I large number of images is required, it would be instead be recommended for the knowledge author to instead product a PowerPoint sequence which can then be attached to the core article instead (via the 'Details' tab, then 'Attachments').
11 Knowledge RTF Editor Copying/Pasting Images
“If I 'Copy and Paste' an image from the Internet into the RTF Editor, what happens?”
Unlike the above method detailed in section 10 of this document, the copy/pasting of images does not store images within the Remedy Database. Instead, a reference to the image is created with a URL linking to the original source. For this reason, this method is not recommended unless you are sure that your users will also have access to the same URL.
For example, if I use Google Images to locate an image, then right-click it and select 'Copy Image':
I then switch to my RTF editor and select 'Paste' which places the copied image into my article:
If I then save the article and view it as a regular user would, the image appears perfectly - however if you inspect the code behind the article whilst viewing it, it is clear that the source is not the Remedy database and instead the image is being retrieved at runtime from the internet.
This method of embedding images is acceptable but only after considering the following:
- Users viewing the article will not see images if they do not have access to the location from which they were copied/pasted, i.e. if they don't have internet access or if an internal corporate firewall blocks access to the URL.
- If you copy/paste from an internet location and the site goes down or the content is changed and the image is removed, this will render the knowledge article with missing images.
- You may not have permission to utilise the image that you have copied/pasted from the internet. Sites track the usage of images and in some circumstances will prosecute if an image is being called from an application/web-page without explicit permission.
In addition, under certain scenarios the following can also occur, causing the images to be unworkable:
- Image is embedded but via a binary BLOB string which is placed into the Knowledge Article content itself, therefore bloating the content of the article, some Web Browsers may also struggle to render this image. In addition images encoded into the article itself as binary bypass the 'Attachment Size Limits' imposed within ARS, this can cause significant database bloat.
- Image is successfully embedded but most often as a BMP instead of a more optimised image format, such as JPG, GIF or PNG.
Based on the above, embedding images properly (using the RTF editor attachment tools) is the recommended method.
12 Knowledge RTF Editor Issues
“When copying/pasting content into the Knowledge RTF Editor, I sometimes struggle with the formatting?”
When copying and pasting content from a source into the RTF editor, certain formatting elements can display oddly for no obvious reason. This is generally due to the source application adding additional and hidden HTML/CSS tags to the content which the RTF editor will attempt to interpret.
To avoid this, the following steps are recommended when migrating content via copy/paste.
- Highlight the content from the source and select 'Copy'.
- Open 'Notepad' or a generic text editor and 'Paste' the content into it.
- Select all of the content again within the above editor and select 'Copy'.
- Switch to the RTF editor and select 'Paste'.
The above process will generally ensure that such hidden content is stripped from the block that has been copied, however this is dependent on the source application from which the 'copy' operation is executed.
13 Knowledge Attachment Indexing
"When attaching files to Knowledge Files, they're automatically indexed by FTS, but what types of content can be indexed?"
FTS is able to index a variety of attachment types, the information indexed depends on the content and file type. the following formats are supported by the 8.1 FTS plugin:
- Hypertext markup language (HTML) format
- XML and derived formats
- Microsoft Office document formats (Word 97 and later--see the note that follows)
- OpenDocument format (OpenOffice 1.0 and later--see the note that follows)
- Portable document format (PDF) (versions 1.0 through 9.0)
- Electronic publication format (digital books)
- Rich text format (RTF)
- Compression and packaging formats (.zip, .tar, .bzip2, .ar, .cpio )
- Text formats (Most Unicode and ISO 8859 documents in plain text)
- Audio formats (extracts Lyrics [if present] and any metadata from MP3, MIDI, and other simple audio formats)
- Image formats (extracts metadata from image formats supported by the java platform)
- Video formats (supports only Flash video format using a simple parsing algorithm)
- Java class files and archives (extracts class names and method signatures from Java class files and the .jar files containing them)
- The mbox format (extracts email messages from the mbox format used by many email archives and UNIX mailboxes)
Regarding attachments, it is important to note that attachment sizes should always be limited (at the ARS level) to avoid end-users attaching large files which will bloat the database size and/or therefore potentially cause performance issues as well as higher bandwidth utilisation.
For more information or for the latest supported attachment list in relation to FTS, please refer to this link:
14 Knowledge Stop/Ignore Words
"Keywords that I search for are not being highlighted or found within Knowledge Results?"
The OOTB FTS application leverages what are known as 'Stop Words', also referred to via the configuration GUI as 'Ignore Words', this feature consists of a semi-colon delimiter separated list of standalone letters and words which the FTS is forced to ignore (not index).
Having a 'Stop Word' list provides three key benefits:
- The overall FTS Index size is smaller.
- Irrelevant words are not indexed and therefore results are more accurate.
- Reindex time is reduced.
The 'Stop Word' configuration can be accessed directly via the file system (within /fts/ftsconfiguration/) or alternatively via the 'AR System Administration -> Server Information' form, then select the 'FTS' tab.
An OOTB set of 'Stop Words' are provided upon installation but these potentially incorporate terms that your business/organisation might consider 'content relevant', therefore you should edit the 'Stop Words' list to remove such terms.
The OOTB set appear as follows:
Many examples of words that might be relevant to an organisation exist, such as "youngest", "highest" and "smallest" and these should therefore be removed on a per-customer/per-implementation basis, but consider such removals very carefully as each removal of a 'Stop Word' will equate to a slightly enlarged FTS index.
15 FTS Fortification with Remedy 9
15.1 "When is FTS Fortification required with Remedy 9 - fresh installation vs upgrade?"
This is a commonly asked question and fortunately the answer is fairly simple. Previous installers (prior to version 9) did not take care of FTS Fortification on your behalf, therefore if upgrading from 8.0 to 8.1 or 7.6.04 to 8.1, you would be expected to run the FTS Fortification process separately after install.
Remedy 9 dramatically simplifies this process but the considerations are different depending on the context, as explained below:
- New Installation of Remedy 9:
Upon fresh installation, Remedy 9 is already fully optimised and FTS Fortified, therefore no additional activities are required - however, future customisations to the system may alter or impact the OOTB FTS strategy, therefore caution should be taken when making changes to the OOTB application.
- Upgraded Installation to Remedy 9:
If upgrading to Remedy 9 from an earlier version of Remedy, the FTS Fortification process is automatically applied during install - all base objects are reset back to an OOTB state with best-practice FTS Fortification attributes applied. However, it should be noted that if Overlays are present on FTS fortified forms, then any customised elements will potentially overrule the best-practice FTS Fortification and therefore the fully optimal state would not be achieved until such Overlays are removed or altered to mirror the underlying FTS attributes.
In such a context, you should evaluate all overlaid objects for that schema, determine which have custom FTS attributes and either remove or reconcile appropriately.
15.2 "I have run FTS Fortification but it didn't make any difference?"
It is common for FTS Fortification to be run twice on an implementation, one run after the next. On occasion, the first scan/sweep of invalid FTS properties doesn't detect all elements, this is not reported in the Fortification report, therefore running the process again generally will generally resolve such issues.
16 Knowledge & FTS Common Issues/Problems
16.1 "The Knowledge Article 'Assignment' menu only shows assignees from my Company?"
The Menu object referenced by the ‘Assignee’ field (RKM:Assignee-RKM-Q) uses two hidden fields as part of its qualification (CurrentAssignee_Company and CurrentAssignee_Organisation), consequently it’ll only show specific assignees when selected. The hidden fields store the 'Company' and 'Organisation' values for the user to whom the article is currently assigned.
This OOTB limitation feels illogical to me and the hope would be that this functionality will be opened up in a future release, as a consequence I would recommend a very simple customisation.
- 1. Overlay the ‘Menu’ object “RKM:Assignee-RKM-Q”.
- 2. Amend the qualification and remove the first two elements:
- “(‘Company’ = $302300585$) AND ‘Support Organisation’ = $302300586) AND”
- 3. Save the Menu and flush the Mid Tier Cache.
Please note that this customisation has not been tested and that you implement it at your own risk. Neither I nor BMC can be held responsible for customisations which should always be thoroughly tested in a development environment first.
16.2 "Users are reporting that the 'Save' button in the Knowledge Article sometimes disappears?"
This issue is intermittant and at this time I am not aware of the actual root cause besides some form of inherent corruption within the creation of cache for the form 'RKM:KnowledgeArticleManager'. The fix for this is to clear the Mid Tier Cache which is not always viable in a Production environment but it does resolve the issue, for a period of time at least, but I urge users to report this issue to BMC Support to fully investigate and to gain resolution for it.
17 Knowledge Tab/Window Issues
"Sometimes a Knowledge Article opens in a new Browser Tab, sometimes in the Breadcrumb, why?"
There is an inconsistency in the product whereby a Knowledge Article will open in a new Web Browser tab but under other circumstances it will open within the same window as the Incident using the 'breadcrumb' system. There are two scenarios which explain the reason for these two different behaviours.
- Create a new Incident or Problem ticket, then click 'Search Knowledge' = Articles open in a New Tab/Window.
- Create a new Incident or Problem ticket, save the ticket, then click 'Search Knowledge' = Articles open in same window via breadcrumb.
The above illustrates that by saving the ticket before prompting a 'Search Knowledge' function, the behaviour of articles opened from the search results window is changed.
For most Service Agents, it is desirable to be able to refer to the Knowledge Article and the ticket (Incident or Problem) at the same time, therefore the 'New Tab' behaviour is preferable. For that reason I intend to investigate and provide details of a minor customisation which can be implemented to make this the default behaviour regardless of the context.
There is a simple customisation which can be implemented to resolve this issue.
18 Knowledge Management with Smart IT
"I am using Remedy Knowledge Management, what are the implications in Smart IT"
Smart IT is a fresh and innovative but early product, therefore process and functionality is continually evolving with each release, therefore the information contained within this section is subject to change. Also note that the version note included within brackets at the end of each bullet is to illustrate the version in which that element has been detected, it may also apply to later versions.
At this time, my observations concerning Knowledge within Smart IT are limited, however, be aware of the following:
- When creating a new Knowledge Article from an Incident, unlike the process in Remedy, no elements from the Incident are passed into the newly created article (Smart IT 1.2/1.3).
- When disabling Knowledge Templates (using the 'Manage Knowledge Sources dialog' in Remedy), the Smart IT Tomcat instance should be restarted to prompt a recache of the templates that are available to the end user (Smart IT 1.2/1.3).
19 Knowledge Article 'Last Modified Date/By' Fields
"Why are the 'Last Modified By' and 'Last Modified Date' fields updated when I haven't edited the article?"
Unfortunately due to the way in which these two 'System' fields work, whenever a 'record' in Remedy is updated, the 'Last Modified By' and 'Last Modified Date' fields are also updated respectively. Due to the way in which the 'Use', 'Relate' and 'Viewed' features work, each time that any is triggered the 'record' (the Knowledge Article) is updated with incremented counters, consequently the 'Last Modified' set of fields are also updated which is confusing because the expectation is that these fields are only ever updated when a change is made to the article.
In summary, these fields are updated on the following actions:
- When the Knowledge Article is opened in 'Advanced' View.
- When the Knowledge Article is 'Used'.
- When the Knowledge Article is 'Related'.
- When the Knowledge Article is 'Viewed'.
- When the Knowledge Article is genuinely edited.
With the Smart IT UX taking the place of the traditional form views, it is unlikely that this behaviour will be changed from the Remedy perspective, with that in mind the following workarounds are recommended:
- Users should be instructed to reference the 'Revisions' tab or the 'Article History' dialog to review the dates/times on which the article was genuinely edited/revised.
- Create two new custom fields ('Last Edited Date' and 'Last Edited By') and a new Filter which will push values into these two custom fields on the click of the 'Save' button on the 'RKM:KnowledgeArticleManager' form. Overlay the existing 'Last Modified' fields and set their visibility to 'Hidden', then position these two new fields in their place.
My recommendation is the second work-around, it is a very simple customisation and will not affect future upgrades.
20 Knowledge Article 'Update Requests'
20.1 "Moving an 'Update Request' to a 'Completed' or 'Rejected' status?"
There is an design decision which is a little illogical relating to Update Requests, this design persists even in v9.1 of ITSM. This specific workflow enforces the addition of a 'Comment' to an 'Update Request' whenever moving the 'Update Request' status to 'Completed' or 'Rejected', however this workflow only triggers if the logged-in user is the assignee of the Knowledge Article, the consequence of which is that users to whom the article is not currently assigned are not shown any warning.
This is illogical, for a user that is not currently working on a Knowledge Article to be able to close an Update Request without any comment or justification is not good practice. I would recommend testing the following customisation (theory, not tested) which would prompt the warning to be shown to all users, regardless of their involvement in the respective Knowledge Article. Edit the Active Link 'RKM:URQ:Save_130_ModifyStatusForceAddComment' and remove the following string from the 'Qualification':
- " AND ('z1D_IsUserAssignee' != $NULL$)"
21 Knowledge Article 'Status Transitions'
"Is it possible to disable the 'Protected' Status Transitions?"
There are a set of Status Transition records which are 'protected' OOTB from any changes being made to them, the protection is enforced via workflow and therefore it is possible to disable the workflow responsible (1 x Filter and 1 x Active Link) therefore allowing you to make changes to them. Whilst I have tested this in 8.x with success but I am yet to assess the impact that such changes would have on Smart IT.
22 Knowledge Process Recommendations
"Managing Knowledge creation, visibility groups and knowledge consumption"
Smart IT is aligned with KCS best-practice for knowledge, for that reason you will find some subtle differences from the old Mid Tier interface, such as the 'All' filter being available whereas in Mid Tier this is only indeed available to Knowledge Admins.
KCS (a global knowledge framework) drives the concept of early adoption of new knowledge content prior to formal publishing and verification, this is to ensure that knowledge articles don't sit for months in a 'draft' state without being pushed forward or forgotten about. It's also to provide content to agents as soon as it is possible, regardless of the format/quality.
Now I don't necessarily agree that this is a good idea (and like ITIL, no 'framework' is a perfect fit for all organization) because content which hasn't been content-checked or formatted correctly can actually cause more confusion than benefit to an agent, it can also introduce clutter to a knowledgebase if numerous duplicated, inconsistent, unformatted articles exist and are suggested to agents.
For the above reasons, I would recommend the following approach:
- Introduce a single global or a set of 'Knowledge Contributor' Support Groups, article creators/contributors in each area of the organization should be added to these respective groups and any new articles should initially be assigned to these groups only from the 'In Progress' status.
- During creation of articles, contributors should be clearly instructed not to apply any Visibility Groups, this will prevent the wider organisation from seeing it before it's ready for consumption. Only the author's respective 'Knowledge Contributor' Support Group (and it's members) will have visibility by default.
- Use the 'SME Review' or 'Content Review' phases for consideration of Visibility Group allocation, after-all it really should be an SME (or content reviewer/support group leader) who determines to whom an article is relevant/visible, not the contributor - this is especially important with sensitive article content which needs consideration of GDPR, privacy and contractual restrictions. By this time, the content has been reviewed and improved, therefore is suitable for a wider audience prior to approval.