[](https://crates.io/crates/ntdsextract2)
- [ntdsextract2](#ntdsextract2)
- [Why do you write a tool that's already there and working?](#why-do-you-write-a-tool-thats-already-there-and-working)
- [Installation](#installation)
- [Usage](#usage)
- [Inspecting the directory information tree visually](#inspecting-the-directory-information-tree-visually)
- [Search for entries](#search-for-entries)
- [Displaying a single entry](#displaying-a-single-entry)
- [Displaying the tree structure of the AD](#displaying-the-tree-structure-of-the-ad)
- [Creating a timeline](#creating-a-timeline)
- [Enumerating ...](#enumerating-)
- [... users](#-users)
- [... groups](#-groups)
- [... computers](#-computers)
- [... types](#-types)
- [Configuring the global timestamp format](#configuring-the-global-timestamp-format)
- [Forensics details](#forensics-details)
- [Interpreting timestamps](#interpreting-timestamps)
# ntdsextract2
<img align="right" width="128px" src="https://raw.githubusercontent.com/janstarke/ntdsextract2/main/doc/images/ntdsextract2.jpeg">
This aims to be a replacement of <https://github.com/csababarta/ntdsxtract/> by @csababarta.
## Why do you write a tool that's already there and working?
1. ntdsxtract is using Python 2.7, which makes it hard to use on modern systems
1. There has been no change since a lot of time (the last commit is from February 2016), which suggests that Csaba has other stuff to do at the moment. That's OK. But Windows *is* changing, and therefore the tools to analyze Windows Systems has to adapt. As I don't like some architectural decisions Csaba has made, I started my own development.
<table>
<tr>
<td>
<a href="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/attributes.png"><img src="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/attributes.png" /></a>
</td>
<td>
<a href="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/acl.png"><img src="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/acl.png" /></a>
</td>
<td>
<a href="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/members.png"><img src="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/members.png" /></a>
</td>
</tr>
</table>
# Installation
```bash
cargo install ntdsextract2
```
# Usage
```
Display contents of Active Directory database files (ntds.dit)
Usage: ntdsextract2 [OPTIONS] <NTDS_FILE> <COMMAND>
Commands:
user Display user accounts
group Display groups
computer display computer accounts
timeline create a timeline (in flow-record format)
types list all defined types
tree display the directory information tree
entry display one single entry from the directory information tree
search search for entries whose values match to some regular expression
acl search for signs of ACL manipulation
walk open the NTDS database in an interactive mode
info show metainfo about the NTDS database
help Print this message or the help of the given subcommand(s)
Arguments:
<NTDS_FILE>
name of the file to analyze
Options:
-v, --verbose...
Increase logging verbosity
-q, --quiet...
Decrease logging verbosity
--use-cache
set this flag if you want to use a cache file, to speed up you work with ntdsextract2
--cache-file <CACHE_FILE>
cache file to use. If you use the cache, but don't specify a filename, ntdsextract2 will generate a filename and store a cache in /tmp
Be aware that ntdsextract2 will overwrite this file at any time without any warning!!!
-h, --help
Print help (see a summary with '-h')
-V, --version
Print version
```
## Inspecting the directory information tree visually
```
Usage: ntdsextract2 <NTDS_FILE> walk
```
opens a screen which looks similar to the following:
<a href="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/attributes.png"><img src="https://codeberg.org/janstarke/ntdsextract2/raw/commit/6ffa4adb3c0165a688cac3a6263ed4e2a7312ee3/doc/images/attributes.png" /></a>
</td>
where you can walk around and inspect the entries that you like. I suggest that you also use the `--use-cache` switch, which can speed up the loading time.
## Search for entries
```
Usage: ntdsextract2 <NTDS_FILE> search [OPTIONS] <REGEX>
Arguments:
<REGEX> regular expression to match against
Options:
-i, --ignore-case case-insensitive search (ignore case)
-v, --verbose... Increase logging verbosity
-q, --quiet... Decrease logging verbosity
-h, --help Print help
```
## Displaying a single entry
```
Usage: ntdsextract2 <NTDS_FILE> entry [OPTIONS] <ENTRY_ID>
Arguments:
<ENTRY_ID>
id of the entry to show
Options:
--sid
search for SID instead for NTDS.DIT entry id. <ENTRY_ID> will be interpreted as RID, wich is the last part of the SID; e.g. 500 will return the Administrator account
-F, --format <ENTRY_FORMAT>
[default: simple]
Possible values:
- json: use JSON format
- table: display a formatted table
- simple: use a simple key-values based format
-v, --verbose...
Increase logging verbosity
-q, --quiet...
Decrease logging verbosity
-h, --help
Print help (see a summary with '-h')
```
## Displaying the tree structure of the AD
```
Usage: ntdsextract2 <NTDS_FILE> tree [OPTIONS]
Options:
--max-depth <MAX_DEPTH> maximum recursion depth [default: 4]
-v, --verbose... Increase logging verbosity
-q, --quiet... Decrease logging verbosity
-h, --help Print help
```
## Creating a timeline
```
Usage: ntdsextract2 <NTDS_FILE> timeline [OPTIONS]
Options:
--all-objects show objects of any type (this might be a lot)
--include-deleted include also deleted objects (which don't have an AttObjectCategory attribute)
-v, --verbose... Increase logging verbosity
-q, --quiet... Decrease logging verbosity
-h, --help Print help
```
## Enumerating ...
### ... users
```
Usage: ntdsextract2 <NTDS_FILE> user [OPTIONS]
Options:
-F, --format <FORMAT>
Output format
[default: csv]
[possible values: csv, json, json-lines]
-A, --show-all
show all non-empty values. This option is ignored when CSV-Output is selected
-D, --include-dn
include the distinguished name (DN) in the output.
Note that this property is not an attribute of the AD entry iself; instead it is constructed from the relative DN (RDN) of the entry and all of its parents. That's why this property is normally not shown.
--member-of <MEMBER_OF_ATTRIBUTE>
specify which attribute shall be used to display group memberships
[default: rdn]
Possible values:
- sid: show the Security ID (SID)
- rdn: show the relative distinguished name (RDN) value
- dn: show the distinguished name (DN)
- sam: show the samAccountName attribute
-v, --verbose...
Increase logging verbosity
-q, --quiet...
Decrease logging verbosity
-h, --help
Print help (see a summary with '-h')
```
### ... groups
```
Usage: ntdsextract2 <NTDS_FILE> group [OPTIONS]
Options:
-F, --format <FORMAT>
Output format
[default: csv]
[possible values: csv, json, json-lines]
-A, --show-all
show all non-empty values. This option is ignored when CSV-Output is selected
-D, --include-dn
include the distinguished name (DN) in the output.
Note that this property is not an attribute of the AD entry iself; instead it is constructed from the relative DN (RDN) of the entry and all of its parents. That's why this property is normally not shown.
--member-of <MEMBER_OF_ATTRIBUTE>
specify which attribute shall be used to display group memberships
[default: rdn]
Possible values:
- sid: show the Security ID (SID)
- rdn: show the relative distinguished name (RDN) value
- dn: show the distinguished name (DN)
- sam: show the samAccountName attribute
-v, --verbose...
Increase logging verbosity
-q, --quiet...
Decrease logging verbosity
-h, --help
Print help (see a summary with '-h')
```
### ... computers
```
Usage: ntdsextract2 <NTDS_FILE> computer [OPTIONS]
Options:
-F, --format <FORMAT>
Output format
[default: csv]
[possible values: csv, json, json-lines]
-A, --show-all
show all non-empty values. This option is ignored when CSV-Output is selected
-D, --include-dn
include the distinguished name (DN) in the output.
Note that this property is not an attribute of the AD entry iself; instead it is constructed from the relative DN (RDN) of the entry and all of its parents. That's why this property is normally not shown.
--member-of <MEMBER_OF_ATTRIBUTE>
specify which attribute shall be used to display group memberships
[default: rdn]
Possible values:
- sid: show the Security ID (SID)
- rdn: show the relative distinguished name (RDN) value
- dn: show the distinguished name (DN)
- sam: show the samAccountName attribute
-v, --verbose...
Increase logging verbosity
-q, --quiet...
Decrease logging verbosity
-h, --help
Print help (see a summary with '-h')
```
### ... types
```
Usage: ntdsextract2 <NTDS_FILE> types [OPTIONS]
Options:
-F, --format <FORMAT> Output format [default: csv] [possible values: csv, json, json-lines]
-v, --verbose... Increase logging verbosity
-q, --quiet... Decrease logging verbosity
-h, --help Print help
```
## Configuring the global timestamp format
Per default, `ntdsextract2` uses an RFC3339-compliant data format. If you want to, you can change the data format
being used by setting the `DFIR_DATE` environment variable. Let's look at an example:
```shell
$ ntdsextract2 tests/data/ntds_plain.dit user -F json-lines |jq 'select (.rdn == "Administrator")'
```
```json
{
"sid": "S-1-5-21-1467604378-2733498025-3532005688-500",
"user_principal_name": null,
"rdn": "Administrator",
"sam_account_name": "Administrator",
"sam_account_type": "SAM_USER_OBJECT",
"user_account_control": "ADS_UF_NORMAL_ACCOUNT | ADS_UF_DONT_EXPIRE_PASSWD",
"logon_count": 4,
"bad_pwd_count": 0,
"admin_count": null,
"is_deleted": false,
"primary_group_id": 513,
"primary_group": "Domänen-Benutzer",
"member_of": [
"Richtlinien-Ersteller-Besitzer",
"Schema-Admins",
"Administratoren",
"Organisations-Admins",
"Domänen-Admins"
],
"comment": null,
"record_time": "2023-11-15T06:33:44+0000",
"when_created": "2023-11-15T06:33:44+0000",
"when_changed": "2023-11-15T06:41:50+0000",
"last_logon": "2023-11-15T06:41:50+0000",
"last_logon_time_stamp": "2023-11-15T06:41:50+0000",
"account_expires": "+30828-09-14T02:48:05+0000",
"password_last_set": "2023-11-15T05:40:32+0000",
"bad_pwd_time": "1601-01-01T00:00:00+0000"
}
```
```shell
```json
{
"sid": "S-1-5-21-1467604378-2733498025-3532005688-500",
"user_principal_name": null,
"rdn": "Administrator",
"sam_account_name": "Administrator",
"sam_account_type": "SAM_USER_OBJECT",
"user_account_control": "ADS_UF_NORMAL_ACCOUNT | ADS_UF_DONT_EXPIRE_PASSWD",
"logon_count": 4,
"bad_pwd_count": 0,
"admin_count": null,
"is_deleted": false,
"primary_group_id": 513,
"primary_group": "Domänen-Benutzer",
"member_of": [
"Administratoren",
"Schema-Admins",
"Domänen-Admins",
"Organisations-Admins",
"Richtlinien-Ersteller-Besitzer"
],
"comment": null,
"record_time": "2023-11-15 06:33:44 (UTC)",
"when_created": "2023-11-15 06:33:44 (UTC)",
"when_changed": "2023-11-15 06:41:50 (UTC)",
"last_logon": "2023-11-15 06:41:50 (UTC)",
"last_logon_time_stamp": "2023-11-15 06:41:50 (UTC)",
"account_expires": "+30828-09-14 02:48:05 (UTC)",
"password_last_set": "2023-11-15 05:40:32 (UTC)",
"bad_pwd_time": "1601-01-01 00:00:00 (UTC)"
}
```
See the difference?
# Forensics details
## Interpreting timestamps
Active Directory stores its timestamp values mostly as [`JET_coltypCurrency`](https://learn.microsoft.com/en-us/windows/win32/extensible-storage-engine/jet-coltyp), because it is the only supported 64bit fixed integer value. So, in fact, timestamps are stored as unsigned 64bit values. This leaves enough space to use the [`FILETIME`](https://learn.microsoft.com/en-us/office/client-developer/outlook/mapi/filetime) structure.
This structure specifies the number of 100 nanoseconds since January 1, 1601. The minimum value is - of course - *`1601-01-01T00:00:00`* with a binary value of `0x0000000000000000`. The maximum value in the most cases is `0x7FFFFFFFFFFFFFFF`, which corresponds to the value *`30828-09-14T02:48:05.4775807`*. However, it is up to each software to interpret those special values.
The following table shows how each of the AD attributes are to be interpreted:
`record_time`| nothing specific |nothing specific|
`when_created`| nothing specific |nothing specific|
`when_changed`| nothing specific | nothing specific|
[`last_logon`](https://learn.microsoft.com/en-us/windows/win32/adschema/a-lastlogon)| last logon time is unknown | |
[`last_logon_time_stamp`](https://learn.microsoft.com/en-us/windows/win32/adschema/a-lastlogontimestamp)| nothing specific but likely related to above | |
[`account_expires`](https://learn.microsoft.com/en-us/windows/win32/adschema/a-accountexpires)| "If at any point in time an account which was configured with an expiration time is set back to Never Expires, the **accountExpires** attribute is then set to 0." |"When an account is created, the account is initially set to Never Expire. The **accountExpires** attribute is set to the default of *9223372036854775807*, a value which corresponds the maximum value of a 64-bit signed integer." |
`password_last_set`| if UAC attr does not contain `UF_DONT_EXPIRE_PASSWD` then user must change password at next logon| |
[`bad_pwd_time`](https://learn.microsoft.com/en-us/windows/win32/adschema/a-badpasswordtime)| the last time an incorrect password was used is unknown| |