Page History
Overview
iDNA Applications collects design data in order to surface insights and complexity about the various applications in your environment. To determine if there are issues with design collections either
- click Click on the area below STATUS area in the top menu bar
- or Or select it from the menu Settings > Design analysis status.
From here, Here you will receive all messages about why iDNA had a failure collecting design information regarding a particular database. The issues could be wide and varied to actual issues with the application's design structure (this is a Domino error, not an iDNA one). You get a hint about what is broken and you need to fix the design yourself: This could be permission issues, server down, etc.
Often failures are caused by the IBM / HCL DXL Exporter on which the Design Collection in iDNA Application depends. DXL Exporter is deemed unstable, however there is no other way to export the designs of Notes databases. panagenda works around the frequent crashes by having the DXL exporter running in its own Docker container and when freezes are detected from the outside, it is restarted automatically. For advanced commands, you could go to the iDNA application's console, issue a "docker ps -a--" and you can notice a frequently different uptime of the panagenda_dda Docker container compared to all the other ones. However, once the exporter crashed on one database, we cannot automatically rescan it again due to technical constraints of DXL itself. This is how the list of failed databases builds.
...
see a list with basic information about database instances and their design collection scan results. Scan results (column "Design Collection → Status") should be OK for most, but where it is not OK (or status "FAILED"), additional information can be found in the column "Design Collection → Message".
The three most common issues are:
- Security access issues (server access, database access, database access level)
- Network access issues
- DXL Exporter issues
Security and network issues should be relatively easy to solve. After troubleshooting, the databases need to be re-scanned manually using the re-scan options on the same page.
Troubleshooting DXL Exporter Issues
Sometimes, failures are caused by the DXL Exporter task that runs on the IBM/HCL Domino Server. iDNA depends on this separate task to gather design information from databases. It is not uncommon for the DXL Exporter to have problems with what it considers corrupt data in design elements.
Clean Abort
Normally that results in a clean abort of the scan on Domino side with an Error message like this:
Code Block | ||
---|---|---|
| ||
Execution failed;
at [ResourceIdentifier[server=NamedReference[name=server/acme],
database=NamedReference[name=test/database.nsf]]];
exporterLog=[
<?xml version='1.0'?>
<DXLExporterLog>
<error id='7112'>Error occurred during processing of note ID 0x57A</error>
<warning id='7034'>Inconsistent data encountered %s</warning>
<error>DXL exporter operation failed</error>
</DXLExporterLog>
] |
Inconsistencies like that can usually be fixed by opening up the database in question and navigating to the document identified by its note ID in the XML (e.g. 0x57A in the error message above) and simply re-saving it. To find a document by note ID, several 3rd party development tools are available (e.g. Noteman by MartinScott).
DXL Crash
More severe issues during DXL export will lead to a crash of the DXL exporter. Even though it's not caused by iDNA, but the interaction between the DXL task and the Notes database, iDNA helps buffering the impact of such a crash with its multi-process scan architecture based on Docker.
In such a case, an error message would show up in the column "System Status" and provides additional information to the problem. Database compact and fixup with recovery options are recommended to approach solving these issues. If the problem persists, opening a PMR with HCL might be requried.
Initiating a Re-Scan
The best practice is to simply re-scan failed databases. Therefore:
- Click on the Status column to display its column/filter options
- Select the Filter tab
- Deselect OK so that only FAILED databases will be displayed
- Make sure that Filtered results is selected in the Re-scan Database Designs field and click on Start
Note |
---|
Re-scanning databases is a multistep multi-step process, that will take up to 2 days before the newly re-scanned databases will appear in the interface. Do not re-scan all databases unless absolutely necessary as it may run over several days and put significant strain on the appliance and infrastructure. In some cases several re-scans might be required. |
You can make use of further filter options in this view to restrict the re-scan scope further down to DXL Exporter crashes.
...