Commit Graph
19 Commits
Author SHA1 Message Date
Ivan Nardi 5704e4c142 STUN: add detection of ZOOM peer-to-peer flows (#1825)
See: "Enabling Passive Measurement of Zoom Performance in Production Networks"
https://dl.acm.org/doi/pdf/10.1145/3517745.3561414
2022-12-11 23:07:35 +01:00
Ivan Nardi ca5ffc4988 TLS: improve handling of ALPN(s) (#1784)
Tell "Advertised" ALPN list from "Negotiated" ALPN; the former is
extracted from the CH, the latter from the SH.

Add some entries to the known ALPN list.

Fix printing of "TLS Supported Versions" field.
2022-10-25 17:06:29 +02:00
Nardi Ivan 1f345b311f Sizes of LRU caches are now configurable
0 as size value disable the cache.

The diffs in unit tests are due to the fact that some lookups are
performed before the first insert: before this change these lookups
weren't counted because the cache was not yet initialized, now they are.
2022-09-23 18:33:48 +02:00
Ivan Nardi a7c2734b38 Remove classification "by-ip" from protocol stack (#1743)
Basically:
* "classification by-ip" (i.e. `flow->guessed_protocol_id_by_ip` is
NEVER returned in the protocol stack (i.e.
`flow->detected_protocol_stack[]`);
* if the application is interested into such information, it can access
`ndpi_protocol->protocol_by_ip` itself.

There are mainly 4 points in the code that set the "classification
by-ip" in the protocol stack:  the generic `ndpi_set_detected_protocol()`/
`ndpi_detection_giveup()` functions and the HTTP/STUN  dissectors.

In the unit tests output, a print about `ndpi_protocol->protocol_by_ip`
has been added for each flow: the huge diff of this commit is mainly due
to that.

Strictly speaking, this change is NOT an API/ABI breakage, but there are
important differences in the classification results. For examples:
* TLS flows without the initial handshake (or without a matching
SNI/certificate) are simply classified as `TLS`;
* similar for HTTP or QUIC flows;
* DNS flows without a matching request domain are simply classified as
`DNS`; we don't have `DNS/Google` anymore just because the server is
8.8.8.8 (that was an outrageous behaviour...);
* flows previusoly classified only "by-ip" are now classified as
`NDPI_PROTOCOL_UNKNOWN`.

See #1425 for other examples of why adding the "classification by-ip" in
the protocol stack is a bad idea.

Please, note that IPV6 is not supported :(  (long standing issue in nDPI) i.e.
`ndpi_protocol->protocol_by_ip` wil be always `NDPI_PROTOCOL_UNKNOWN` for
IPv6 flows.

Define `NDPI_CONFIDENCE_MATCH_BY_IP` has been removed.

Close #1687
2022-09-20 22:24:47 +02:00
Ivan Nardi 4f584f78a0 Fix ndpi_do_guess() (#1731)
Avoid a double call of `ndpi_guess_host_protocol_id()`.
Some code paths work for ipv4/6 both
Remove some never used code.
2022-09-12 19:28:41 +02:00
Ivan Nardi 0a47f745cc Avoid useless host automa lookup (#1724)
The host automa is used for two tasks:
* protocol sub-classification (obviously);
* DGA evaluation: the idea is that if a domain is present in this
automa, it can't be a DGA, regardless of its format/name.

In most dissectors both checks are executed, i.e. the code is something
like:

```
ndpi_match_host_subprotocol(..., flow->host_server_name, ...);
ndpi_check_dga_name(..., flow->host_server_name,...);

```

In that common case, we can perform only one automa lookup: if we check the
sub-classification before the DGA, we can avoid the second lookup in
the DGA function itself.
2022-09-05 13:59:51 +02:00
Ivan Nardi 405a52ed65 Patricia tree, Ahocarasick automa, LRU cache: add statistics (#1683)
Add (basic) internal stats to the main data structures used by the
library; they might be usefull to check how effective these structures
are.

Add an option to `ndpiReader` to dump them; enabled by default in the
unit tests.
This new option enables/disables dumping of "num dissectors calls"
values, too (see b4cb14ec).
2022-07-29 15:25:00 +02:00
Ivan Nardi 172e698bb8 TINC: avoid processing SYN packets (#1676)
Since e6b332aa, we have proper support for detecting client/server
direction. So Tinc dissector is now able to properly initialize the
cache entry only when needed and not anymore at the SYN time; initializing
that entry for **every** SYN packets was a complete waste of resources.

Since 4896dabb, the various `struct ndpi_call_function_struct`
structures are not more separate objects and therefore comparing them
using only their pointers is bogus: this bug was triggered by this
change because `ndpi_str->callback_buffer_size_tcp_no_payload` is now 0.
2022-07-28 12:39:18 +02:00
Ivan Nardi d8d525fff2 Update the protocol bitmask for some protocols (#1675)
Tcp retransmissions should be ignored.

Remove some unused protocol bitmasks.

Update script to download Whatsapp IP list.
2022-07-27 11:46:45 +02:00
Toni ae2bedce3a Improved Jabber/XMPP detection. (#1661)
Signed-off-by: Toni Uhlig <matzeton@googlemail.com>
2022-07-13 17:55:33 +02:00
Ivan Nardi b4cb14ec19 Keep track of how many dissectors calls we made for each flow (#1657) 2022-07-11 09:47:47 +02:00
havsah e16b6a18be Fix Grease values parsing (#1416)
The check for grease was too broad and filtered some valid values.
In particular, the value 257 was skipped because it matched the previous check.
This has been discovered while parsing tests/pcap/443-firefox.pcap
expected ja3:
771,4865-4867-4866-49195-49199-52393-52392-49196-49200-49162-49161-49171-49172-51-57-47-53-10,0-23-65281-10-11-35-16-5-51-43-13-45-28-21,29-23-24-25-256-257,0
previously generated ja3:
771,4865-4867-4866-49195-49199-52393-52392-49196-49200-49162-49161-49171-49172-51-57-47-53-10,0-23-65281-10-11-35-16-5-51-43-13-45-28-21,29-23-24-25-256,0

Signed-off-by: Patrick Havelange <patrick.havelange_ext@softathome.com>
2022-01-21 12:40:35 +01:00
Ivan Nardi 3a087e951d Add a "confidence" field about the reliability of the classification. (#1395)
As a general rule, the higher the confidence value, the higher the
"reliability/precision" of the classification.

In other words, this new field provides an hint about "how" the flow
classification has been obtained.
For example, the application may want to ignore classification "by-port"
(they are not real DPI classifications, after all) or give a second
glance at flows classified via LRU caches (because of false positives).

Setting only one value for the confidence field is a bit tricky: more
work is probably needed in the next future to tweak/fix/improve the logic.
2022-01-11 15:23:39 +01:00
Ivan Nardi b1e9245d94 ndpiReader: slight simplificaton of the output (#1378) 2021-11-27 17:32:23 +01:00
Ivan Nardi b6d9536533 Fixed cleartext protocol assignment (#1357) 2021-10-25 15:04:04 +02:00
Luca Deri 9e97d20c25 Refreshed results list 2021-10-16 12:03:16 +02:00
Luca Deri e8455236bd Updated output 2021-08-07 17:38:33 +02:00
Ivan Nardi cccf794265 ndpiReader: add statistics about nDPI performance (#1240)
The goal is to have a (roughly) idea about how many packets nDPI needs
to properly classify a flow.

Log this information (and guessed flows number too) during unit tests,
to keep track of improvements/regressions across commits.
2021-07-13 12:28:39 +02:00
Luca Deri ac1eaca8a6 Added browser TLS heuristic 2021-05-13 20:00:27 +02:00