Longstanding IBM Notes Database Unread ID Table Replication Issues, Now Solved!

In talking to hundreds of IBM Domino administrators and developers every week over a good many years now, we’ve frequently heard of problems that revolve around unread marks, their structure, and their replication (or lack thereof). In this article, we’ll examine how the very mechanics behind unread marks could be the reason why they are likely to cause issues, and then we’ll take a look at how easy it is to sort these problems out using scanEZ‘s unique, and easy, unread marks management process.

Every Notes database that doesn’t have the ‘Don’t maintain unread marks’ option enabled—essentially an “opt-out” from unread marks—maintains an Unread ID table for each active user, and this table keeps a list containing the NoteIDs of each unread document. When a database is opened, the Unread Marks ID table is transferred to the local file; all updates are performed locally during the current session in which the database is opened.

Provided that the ‘Replicate Unread Marks’ database property is enabled, the Unread ID table does replicate, and is thus supposed to be identical across replicas.

Note: Unread marks were originally designed for Mail databases, and their replication can significantly slow database performance. For optimal performance, consider reviewing both the ‘Don’t maintain Unread Marks’ and ‘Replicate Unread Marks’ properties across your all of your databases by using databaseEZ.

Replicating the Unread Marks ID table entails first a conversion of NoteIDs to UNIDs in order to locate the appropriate documents in the other replica, and then a second conversion of those UNIDs into NoteIDs in order to update the foreign replica’s NoteID table (see fig. 1). This is all for the sake of space economy; NoteIDs are much smaller (4 bytes) than UNIDs (16 bytes).

Thus, Unread ID table discrepancies likely arise from a combination of circumstances: the two-step conversion, the fact that the tables aren’t updated live during a current session, and occurrences of interrupted sessions, client crashes, etc. mix to provide just enough of an opening for errors to occur.

When using the native IBM Notes or Domino Administrator clients, a user is only allowed to view/manage their own Unread ID table. This makes it impossible for administrators to verify (much less remedy) unread marks discrepancy issues reported by users.

In contrast to the native IBM Notes toolkit, scanEZ allows administrators to view AND manipulate not only their own, but also any user’s Unread Marks table in any given database. This unique feature, first introduced in EZSuite V12, finally gives administrators the means to debug and resolve any issues pertaining to unread marks. And using scanEZ’s capabilities, they can easily perform a “manual synchronization” of the state of unread marks across replicas.

Here is how to do it:

Picture this all-too-common scenario: a user calls in and complains that while their replica on server A has all the correct documents marked as unread, it won’t properly replicate this information to their replica on server B.

Using scanEZ, you can sync those databases in 5 steps or less:

Note: For the sake of brevity, and to illustrate a linear process, the method shown here will use Checkbox Selections. The flexibility of the tools allow you to take a few different paths to arrive at the same results.

1) Find the correct documents that are currently unread for a given user

Open scanEZ on the user’s mailbox on server A, select the option Search for Unread< Documents/strong> from the Search menu, and then select the user’s name. Once the unread documents have been found, create a new My Selection virtual folder containing these documents (see fig. 2).

2.1) If your new My Selection folder contains up to—but not more than—990 documents, checkbox select your My Selection folder, and choose the Copy UNID(s) to Clipboard option from the Checkbox Selection menu (see fig. 3). Now you have all of the currently unread documents’ UNIDs on your clipboard. If you have more than 990 documents, skip directly to step 2.2.

Please proceed to step 3.1.

2.2)If your resulting My Selection virtual folder contains a very large amount (many thousands) of documents, you won’t be able to work with UNIDs, unfortunately, as scanEZ’s Search by UNID function currently supports search input of a maximum of 990 UNIDs at a time (contrast this to the IBM Domino Administrator client, which only allows for searching one UNID). But don’t worry; all you’ll need to do is checkbox select your My Selection folder’s contents, and then select the ‘Add to Folder’ option from the Checkbox Selection menu. Create a new temporary folder called “TmpDocSelect” and add all relevant documents to it (see fig. 4). This is a real folder, as opposed to a virtual My Selection folder. Let this folder replicate over to server B.

Please proceed to step 3.2.

3.1) With the document UNIDs on your clipboard, open scanEZ on the database’s replica on server B and select Search by UNID in the Search menu. Paste all UNIDs from your clipboard into the input field and create a My Selection folder called “All Unread”. Next, checkbox select all documents by putting a check mark next to the Documents tree entry and add them to a new My Selection folder called “All Documents” using the Add to > New ‘My Selection’ option from the Checkbox Selection menu (see fig. 5).

Proceed to step 4.

3.2) Once your newly created folder “TmpDocSelect” has replicated over to Server B, open scanEZ on the database’s replica on server B, and use the Quick Retrieve Contents of a View or Folder feature found in the Search menu to select the “TmpDocSelect” folder. Specify the name “All Unread” for the new My Selection to which all of these documents will be added. Checkbox select all documents by putting a check mark next to the entry Documents in the tree, and then add them to a new My Selection folder called “All Documents” using the Add to > New ‘My Selection’ option from the Checkbox Selection menu.

Proceed to step 4.

4) After making sure nothing is selected in the tree, go ahead and checkbox select your “All Unread” My Selection folder. Use the Remove from > Another My Selection > All Documents option path from the Checkbox Selection menu to remove all of these unread documents from the “All Documents” My Selection folder (see fig. 8). Essentially, this will leave all read documents in the “All Documents” My Selection folder.

5) The “All Documents” My Selection folder should now contain all of the read documents. Now, the only thing left to do is to make them officially read for the user. This can be done by simply highlighting the “All Documents” My Selection folder, and selecting the Document menu’s Mark as Read function. Select the user’s name and click OK (see fig. 9). The Unread Marks table has now been manually synchronized.

Replication issues have long plagued Unread ID tables and, consequently, annoyed administrators. We have seen the probable behind-the-scenes reasons why mix-ups can happen, but that comes as cold comfort to those who still need to approach these problems in the traditional method. scanEZ‘s easy unread marks management process gives administrators the power to finally remedy these issues across replicas, as well as providing much-needed time savings and gains in accuracy.