This page is for the IT and security reviewers who approve iDNA Applications before or during an evaluation. It describes what iDNA reads from your HCL Notes and Domino environment, which access it needs for that, and which connections it uses. It applies to every license; the differences on a trial license are in one section at the end. It describes iDNA Applications 3.5.0 and later.
What happens to the data after it has been read (storage, processing, who sees what, and what leaves your network) is described on iDNA Applications Data Flow in Detail.
Overview
iDNA Applications runs on a Linux host in your network (the appliance). It reads from the Domino servers you select, with one Notes ID that you provide: the collection ID.
In short
- Read-only towards Domino. iDNA creates, changes or deletes no documents, designs or ACLs in your databases and runs no agents. The only console command it sends is
show config *, which lists the server's notes.ini settings. - Connections start at the appliance: Notes RPC (TCP 1352) to your Domino servers; HTTP or HTTPS to a Domino web server only if web logs are read as text files.
- From system databases: the Domino Directory, server documents and notes.ini settings,
log.nsf,catalog.nsfanddomlog.nsf; the DLAU and MarvelClient Analyze databases when those optional packages are used. - From application databases: the design and source code of focus applications and templates, the database's record of user activity, profile documents in full, and document metadata. No other document content, and no attachment contents.
- Personal data: Notes user names, directory data about people and groups, IP addresses and client details from web logs and MarvelClient, mail addresses from the router's log. Stored on the appliance and pseudonymized in the analyses by default.
The Collection ID and the Access It Needs
iDNA reads everything with the collection ID. It needs the following access; the reference list is in System Requirements, Access and Permissions.
| Access | Where | Used for |
|---|---|---|
| Access to the server (cross-certified, allowed to access the server) | Every Domino server in scope | Any collection from that server |
| Reader | The Domino Directory (names.nsf) on one server per domain, and by default the directories configured in Directory Assistance | The Domino Directory package |
| Reader | log.nsf and catalog.nsf of every server; domlog.nsf where web logging to it is enabled | The Server Basics, Database Catalog, Notes Client Activity and Web User Activity packages; the optional Domino Event Log and Mail Routing Activity packages |
| Full Remote Console Administrator | Server document of every server in scope | The console command show config * and the console part of the server access check |
| Designer, or Full Access Administration | Designer in the ACL of every focus database and template; or Full Access Administration in the server document | Design analysis. Where the ID has Full Access Administration, iDNA uses it by default. Reader access is the minimum for a database to be opened at all. |
| Reader | The DLAU result database; the MarvelClient Analyze database | The optional HCL DLAU and MarvelClient Analyze packages |
| Reader, or Full Access Administration where the ID has it | Each database a content scan is started for | Content analysis |
- Full Access Administration lets the ID open every database on that server whatever its ACL says, mail files included. iDNA uses it to read designs. Designer access per database is the narrower alternative.
- Servers of other organisations: iDNA collects only from servers certified by an organisation named in your license. Other servers it finds are listed as unlicensed and not collected.
- Server prerequisites (the Catalog and Statlog tasks, session logging, the web log format) are listed in System Requirements.
- The ID file and its password are stored on the appliance so that iDNA can collect without anyone logged in; see Data Flow in Detail, Storage on the Appliance.
Connections
| From → To | Protocol, port | Purpose | When |
|---|---|---|---|
| Appliance → Domino servers | Notes RPC, TCP 1352 | All collection from Domino | Always |
| Appliance → Domino web servers | HTTP or HTTPS, the port you configure | Reading web logs as text files | Only if the Web User Activity package is set up to read text web logs over HTTP |
| Appliance → your DNS servers | DNS | Resolving the Domino servers' host names | Always |
Your Domino servers open no connection to the appliance. Browser access to iDNA, logins through LDAP, email, AI clients and every connection that leaves your network are described in Data Flow in Detail, Who Sees What and What Leaves Your Network.
What iDNA Reads, by Source
iDNA collects in three ways:
- Content packages, set up per server in the admin interface (see Admin Interface and Config Manager). Each package reads one kind of data from the servers it is active on.
- Design analysis of focus databases and templates, with the activity record, profile documents and content age of each focus database.
- Content analysis, started by an iDNA user for a single database.
Content Packages
| Package | Read from | What is read | Personal data |
|---|---|---|---|
| Domino Directory | names.nsf on one server per domain; directories in Directory Assistance | Person documents, group documents (with members and owners), mail-in database documents | Names, organisational data and mail and internet addresses of people; group memberships |
| Server Basics | Every server in scope | Server documents, connection documents, notes.ini settings (through show config *), server statistics (hourly) | Administrator names in server documents |
| Database Catalog | Every server in scope; catalog.nsf, log.nsf | The list of databases with their properties (title, file path, replica ID, template, size, dates), catalog entries, the number of ACL entries per access level, database activity counts | None: ACL entries are counted, not stored by name |
| Notes Client Activity | log.nsf | Session records: who opened which database on which server, when, with read and write counts (every 2 hours) | Notes user names |
| Web User Activity | domlog.nsf, or text web logs over HTTP or HTTPS | Web requests: user name, client address and host name, user agent, requested URL, time. Cookies and referrers are not stored. | User names, IP addresses, host names |
| Domino Event Log (optional) | log.nsf | Events and security events: time, type, severity, server task and message text | Names that appear in event messages |
| Mail Routing Activity (optional) | log.nsf | The router's transfer and delivery events: sender, recipients, size, time | Mail addresses |
| HCL DLAU (optional) | The DLAU result database | Its documents in full, rich text included | User names and the licensing data DLAU recorded about them |
| MarvelClient Analyze (optional) | The MarvelClient Analyze database | Client records: Notes user name, home server, Notes and MarvelClient versions, operating system, operating system user name, IP address | All of these |
Design Analysis
Design analysis covers the focus databases (the database types Application and Mail-in, as classified from the catalog; you can adjust the classification with recategorization rules) and one replica of each template. It runs one database at a time and re-reads a design when it has changed, by default weekly on Saturday.
| Source | Read from | What is read | Personal data |
|---|---|---|---|
| Design | Focus databases and templates | The full design, exported as DXL: every design element and its code (LotusScript, Formula language, JavaScript, Java, XPages), names, comments, signers, last modifiers and dates; the database properties; the ACL with its entries, roles and log. The export is kept on the appliance as a compressed file. No data documents. | Names of signers, modifiers and ACL entries; anything developers wrote into the code, such as names, addresses or credentials |
| User activity record | Focus databases and templates | The activity that Notes records in each database: user names, dates, reads and writes | Notes user names |
| Profile documents | Focus databases | Every profile document in full: all items and their values, its modifiers and revision dates | Whatever the application stores in its profile documents; modifier names |
| Content age | Focus databases (content-age module) | The number of documents and their creation and modification dates, stored as counts per month. No item values, no form names. | None |
More on content age: Content Age and Data Retention Analysis in iDNA Applications.
Content Analysis
An iDNA user starts a content scan for one database at a time; each scan is counted against the license. The scan reads every document's metadata: form name, dates, modifiers, signed and encrypted flags, the names of encryption keys, whether reader and author fields are present, attachment names, sizes and dates, document links, and counts of rich-text elements. iDNA stores aggregates of these (attachments by type and size, modifiers with their counts). It reads no item values and no attachment contents.
Is Document or Mail Content Read?
- Application documents: metadata only, except profile documents, which are read in full.
- Mail files: listed in the database inventory and seen in the session records. Design analysis, profile documents and content age do not open them, because mail files are not focus databases.
- Mail messages: Mail Routing Activity reads the router's log entries and takes sender, recipients, size and time from them; no subjects and no message bodies.
- Rich text and attachments: not read, with two exceptions: DLAU documents are read in full, and content analysis reads attachment names, sizes and dates.
- Source code: read and stored in full, including whatever developers wrote into it.
Personal Data
| Personal data | Read from |
|---|---|
| Notes user names | Session records, web logs, events, the user activity records of databases, ACLs, design signers and modifiers, profile documents, DLAU and MarvelClient records |
| Data about people and groups | Person documents and group documents in the Domino Directory |
| IP addresses and client details | Web logs (client address, host name, user agent, requested URL); MarvelClient records (IP address, operating system user name) |
| Mail addresses | The router's log (Mail Routing Activity); the Domino Directory |
| Values written by applications or developers | Profile documents, event messages, source code |
All collected data is stored on the appliance with real names. User names are pseudonymized in the analyses by default. Which analyses show real names on purpose, who sees what, and what the AI components can reach are described on iDNA Applications Data Flow in Detail.
On a Trial License
A trial license uses the same sources, with the same access, as a full license: the content packages, and the design of every focus database and template, exported and kept as described above. The differences:
- Source code: depending on license options, source code from a subset of databases is analyzed.
- Profile documents and content age are read only for the trial sample.
- The trial sample is a set of up to 50 focus database instances that iDNA selects itself from the middle of the pack in usage and complexity. It is recomputed every night, so it can change while usage data build up. It cannot be chosen.
- Content analysis works as on a full license.
What a trial shows, and where it masks names, is described on iDNA Applications Trial License.
Related Topics
- iDNA Applications Data Flow in Detail: storage, processing, who sees what, what leaves your network, AI clients.
- System Requirements: the access and server settings iDNA needs.
- Admin Interface and Config Manager: where the collection is set up.
- Content Age and Data Retention Analysis in iDNA Applications: what content age measures.
- iDNA Applications License Types and Options: license types and options.
- iDNA Applications Trial License: everything about a trial.
