Cbapi

Latest version: v1.7.10

Safety actively analyzes 638773 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 6 of 6

1.3.2

--------------------------------------

This release introduces the Policy API for CB Defense. A sample ``policy_operations.py`` script is now included
in the ``examples`` directory for CB Defense.

Other changes:

* CB Response

* Bugfixes to the ``User`` Model Object.
* New ``user_operations.py`` example script to manage users & teams.
* Additional ``Team`` Model Object to add/remove/modify user teams.
* New ``check_datasharing.py`` example script to check if third party data sharing is enabled for binaries on any sensor groups.
* Documentation fix for the ``User`` Model Object.
* Fix to the ``watchlist_operations.py`` example script.

1.3.1

-------------------------------------

This is a bugfix release with minor changes:

* CB Response

* Add ``partition_operations.py`` script to demonstrate the use of the StoragePartition model object.
* Fix errors when accessing the ``.start`` attribute of child processes.
* Fix errors generated by the ``walk_children.py`` example script. The output has been changed as well to indicate
the process lifetime, console UI link, and command lines.
* Add an ``.end`` attribute to the Process model object. This attribute reports back either ``None`` if the
process is still executing, or the last event time associated with the process if it has exited. See the
``walk_children.py`` script for an example of how to calculate process lifetime.
* Fix errors when using the ``.parents`` attribute of a Process.
* Add ``wait_for_completion`` flag to ``create_process`` Live Response method, and default to ``True``. The
``create_process`` method will now wait for the target process to complete before returning.

* CB Defense

* Add ``wait_for_completion`` flag to ``create_process`` Live Response method, and default to ``True``. The
``create_process`` method will now wait for the target process to complete before returning.

1.3.0

------------------------------------

This release introduces the Live Response API for CB Defense. A sample ``cblr_cli.py`` script is now included in the
``examples`` directory for both CB Response and CB Defense.

Other changes:

* CB Protection

* You can now create new ``FileRule`` and ``Policy`` model objects in cbapi.

* CB Response

* Added ``watchlist_exporter.py`` and ``watchlist_importer.py`` scripts to the CB Response examples directory.
These scripts allow you to export Watchlist data in a human- and machine-readable JSON format and then re-import them into another CB Response server.
* The ``Sensor`` Model Object now uses the non-paginated (v1) API by default. This fixes any issues encountered when
iterating over all the sensors and receiving duplicate and/or missing sensors.
* Fix off-by-one error in ``CbCrossProcess`` object.
* Fix issue iterating through ``Process`` Model Objects when accessing processes generated from a 5.2 server
after upgrading to 6.1.
* Reduce number of API requests required when accessing sibling information (parents, children, and siblings) from the
``Process`` Model Object.
* Retrieve all events for a process when using ``segment`` ID of zero on a CB Response 6.1 server.
* Behavior of ``Process.children`` attribute has changed:

* Only one entry is present per child (before there were up to two; one for the spawn event, one for the
terminate event)
* The timestamp is derived from the start time of the process, not the timestamp from the spawn event.
the two timestamps will be off by a few microseconds.
* The old behavior is still available by using the ``Process.childprocs`` attribute instead. This incurs a
performance penalty as another API call will have to be made to collect the childproc information.

* ``Binary`` Model Object now returns False for ``.is_signed`` attribute if it is set to ``(Unknown)``.

* Moved the ``six`` Python module into cbapi and removed the external dependency.

1.2.0

------------------------------------

This release introduces compatibility with our new product, CB Defense, as well as adding new Model Objects introduced
in the CB Protection 8.0 APIs.

Other changes:

* CB Response

* New method ``synchronize()`` added to the ``Feed`` Model Object

* Bug fixes and documentation improvements

1.1.1

-----------------------------------

This release includes compatibility fixes for CB Response 6.1. Changes from 1.0.1 include:

* Substantial changes to the ``Process`` Model Object for CB Response 6.1. See details below.
* New ``StoragePartition`` Model Object to control Solr core loading/unloading in CB Response 6.1.
* New ``IngressFilter`` Model Object to control ingress filter settings in CB Response 6.1.
* Fix issues with ``event_export.py`` example script.
* Add ``.all_events`` property to the ``Process`` Model Object to expose a list of all events across all segments.
* Add example script to perform auto-banning based on watchlist hits from CB Event Forwarder S3 output files.
* Add bulk operations to the ``ThreatReport`` and ``Alert`` Query objects:

* You can now call ``.set_ignored()``, ``.assign()``, and ``.change_status()`` on an ``Alert`` Query object to change
the respective fields for every Alert that matches the query.
* You can now call ``.set_ignored()`` on a ``ThreatReport`` Query object to set or clear the ignored flag for every
ThreatReport that matches the query.

Changes to ``Process`` Model Object for CB Response 6.1
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

CB Response 6.1 uses a new way of recording process events that greatly increases the speed and scale of collection,
allowing you to store and search data for more endpoints on the same hardware. Details on the new database format
can be found on the Developer Network website at the `Process API Changes for CB Response 6.0
<https://developer.carbonblack.com/reference/enterprise-response/6.1/process-api-changes/>`_ page.

The ``Process`` Model Object traditionally referred to a single "segment" of events in the CB Response database. In
CB Response versions prior to 6.0, a single segment will include up to 10,000 individual endpoint events, enough to
handle over 95% of the typical event activity for a given process. Therefore, even though a ``Process`` Model Object
technically refers to a single *segment* in a process, since most processes had less than 10,000 events and therefore
were only comprised of a single segment, this distinction wasn't necessary.

However, now that processes are split across many segments, a better way of handling this is necessary. Therefore,
CB Response 6.0 introduces the new ``.group_by()`` method. This method is new in cbapi 1.1.0 and is part of five
new query filters available when communicating with a CB Response 6.1 server. These filters are accessible via methods
on the ``Process`` Query object. These new methods are:

* ``.group_by()`` - Group the result set by a field in the response. Typically you will want to group by ``id``, which
will ensure that the result set only has one result per *process* rather than one result per *event segment*. For
more information on processes, process segments, and how segments are stored in CB Response 6.0, see the
`Process API Changes for CB Response 6.0 <https://developer.carbonblack.com/reference/enterprise-response/6.1/process-api-changes/>`_
page on the Developer Network website.
* ``.min_last_update()`` - Only return processes that have events after a given date/time stamp (relative to the
individual sensor's clock)
* ``.max_last_update()`` - Only return processes that have events before a given date/time stamp (relative to the
individual sensor's clock)
* ``.min_last_server_update()`` - Only return processes that have events after a given date/time stamp (relative to the
CB Response server's clock)
* ``.max_last_server_update()`` - Only return processes that have events before a given date/time stamp (relative to the
CB Response server's clock)

Examples for new Filters
~~~~~~~~~~~~~~~~~~~~~~~~

Let's take a look at an example::

>>> from datetime import datetime, timedelta
>>> yesterday = datetime.utcnow() - timedelta(days=1) Get "yesterday" in GMT
>>> for proc in c.select(Process).where("process_name:cmd.exe").min_last_update(yesterday):
... print proc.id, proc.segment
DEBUG:cbapi.connection:HTTP GET /api/v1/process?cb.min_last_update=2017-05-21T18%3A41%3A58Z&cb.urlver=1&facet=false&q=process_name%3Acmd.exe&rows=100&sort=last_update+desc&start=0 took 2.164s (response 200)
00000001-0000-0e48-01d2-c2a397f4cfe0 1495465643405
00000001-0000-0e48-01d2-c2a397f4cfe0 1495465407157
00000001-0000-0e48-01d2-c2a397f4cfe0 1495463680155
00000001-0000-0e48-01d2-c2a397f4cfe0 1495463807694
00000001-0000-0e48-01d2-c2a397f4cfe0 1495463543944
00000001-0000-0e48-01d2-c2a397f4cfe0 1495463176570
00000001-0000-0e48-01d2-c2a397f4cfe0 1495463243492

Notice that the "same" process ID is returned seven times, but with seven different segment IDs. CB Response will
return *every* process event segment that matches a given query, in this case, any event segment that contains the
process command name ``cmd.exe``.

That is, however, most likely not what you wanted. Instead, you'd like a list of the *unique* processes associated with
the command name ``cmd.exe``. Just add the ``.group_by("id")`` filter to your query::

>>> for proc in c.select(Process).where("process_name:cmd.exe").min_last_update(yesterday).group_by("id"):
... print proc.id, proc.segment
DEBUG:cbapi.connection:HTTP GET /api/v1/process?cb.group=id&cb.min_last_update=2017-05-21T18%3A41%3A58Z&cb.urlver=1&facet=false&q=process_name%3Acmd.exe&rows=100&sort=last_update+desc&start=0 took 2.163s (response 200)
00000001-0000-0e48-01d2-c2a397f4cfe0 1495465643405

Page 6 of 6

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.