Search In Packet Logs

You can now search for arbitrary strings in the historical packet logs directly. The only requirements for this search is at least 1 IP address in addition to the search string.

For example in the search below we are looking for the IP address in any packet either sent or received by the host The search is also restricted to an hour worth of packets on 5/7/2018.


So why would you look for an IP address string in the packets? Well, this is normally done when there is more than one proxy and the system is not able to properly identify the proxy chain. In that case the offending IP will be recorded in the x-forwarded-for field of the http headers. Once you find the headers, you can find the real flows and then search again to get the data exchanged specifying the source and destination ports.

But this search feature is much more powerful than that; in fact you can also look retroactively in your packet history using full PERL regular expressions!

If you reached this far in this post, and you are an expert user, you will be wondering about the example above. The search string above would actually match more than because the dots really mean any character (for example 139a182b44c203 would also match). To be more precise you would need to enter:


But suppose you wanted to match a specific set of IP addresses

Using a regular expression you could search for:


Just imagine what you could search for when you are hunting down specific strings or patterns. So, this little new feature (also available through the CLI interface as the option -Q) should really expand the power of our historical packet logging system. It will let you easily dig in your network history for hidden clues of what happened in the past.

Websockets in server mode

You can now use Websockets and disable flash for good. We added support for Websockets instead of Flash for direct real time connection to our sensors to retrieve real time data and payloads. Sensors can be configured in client mode or server mode. In client mode sensors and browser both connect outbound as clients and get connected at the network level by our forwarder. Websockets have been working in this mode for quite sometime. Sensors can also be configured in server mode where the browser needs to connect to the sensor which acts a websocket server. Since sensors only have a self-signed SSL certificate, secure Websockets would not work. To get around this, we added a small (probably the smallest web server you will ever see) that only serves a root certificate authority to be installed in your browser. Go to:

https://<yours sensor ip> for non-Windows systems

https://your sensor ip>:81 for Windows systems

and install the certificate.

Then configure your sensor as a server and make sure to add its static IP address in the sensor configuration. After restarting the sensor, you will be able to connect in server mode using Websockets.

Proxy Detection Support

This has been awaited for some time. The MetaFlows Security System now detects proxied connections. The original IP is swapped with the proxy IP so that it can be correctly identified in the events. This has a dramatic effect on correlation since most proxied hosts only proxy http and use their real IP for DNS and other communications. Using the real IP for correlation and analysis will correctly correlate IDS http events and file downloads with IDS events and service discoveries triggered by different protocols.

When a proxied host is detected, a message of the forms xff=<realip><-<proxyip> is appended to the event and the proxy IP is replaced. So, you will see the real IP not the proxy.

When you analyze the packets data, the system automatically switches back to the proxy IP to look for the packets containing the proxy IP rather than the real IP (since the packets are stored before the IP is replaced).

Here is a real example of two events related to (the real IP) downloading suspicious content through the proxy


Notice that when we detect a proxy a P is associated with the proxied host.

This feature will be available as soon as your sensor is restarted or self-updates. Let us know if you have questions.

Happy hunting!

Your dedicated MetaFlows Team.


Splunk App

We have developed a Splunk network security app available at


It receives events generated by the MetaFlows sensors and breaks them down by the following types:

  • Multisession Analysis
  • High Priority Events
  • IDS Events
  • Network Logs (3rd party logs sent to the sensors)
  • File Transmission Analysis
  • User Discovery
  • Service Discovery
  • Host Discovery
  • Mac Discovery
  • Suspicious URL Transmission Analysis
  • IPS Notifications
  • User Rankings
  • Modsecurity

From the app you can either drill down on Splunk itself or jump to the MetaFlows console to gather more forensic information like packet payloads.

You can install the app by using the Splunk application management tools. In order to send event to Splunk you need to add a configuration line in your /nsm/etc/ startup script of your sensors. The SSL-encrypted syslog messages are sent to the MetaFlows Splunk App through TCP port 3015 (please make sure you sensor can communicate on this port).

It is a early beat version, please let us know how you like it.

Please see more details at

Happy Hunting!

The MetaFlows Team.

WannaCry Ransomware Advisory

It has been all over the news this weekend, a surge in Ransomware under the name ‘wannacry’ that has the potential to cripple large portions of networks due to the way that it spreads.

This is a pretty stealthy piece of malware at the network level, little to no CnC has been confirmed, but at an individual level it doesn’t behave much differently from any other Ransomware that we have seen in the past.

What distinguishes WannaCry is that it has a secondary infection vector that prior Ransomware variants lacked. Like any other, the primary infection vector appears to occur via email attachment (zipped javascript). However, once a machine is compromised it begins to behave more like a worm, able to exploit SMB (windows file sharing) on any systems that it can reach in order to spread its self.

This worm like behavior makes it particularly dangerous. While usually* smb (port 445) is not accessible from the outside world, it is often completely unrestricted within a local network, allowing one infected machine to spread the Ransomware across an entire site.

* This is your reminder to do double check firewall rules and run some external scans to make absolutely certain your windows file shares are not reachable from the outside world.


The following signatures are currently indicators to look out for:

2024218: ET EXPLOIT Possible ETERNALBLUE MS17 Echo Response
2024291: ET TROJAN Possible WannaCry DNS Lookup (trojan.rules)
2024292: ET INFO Bitcoin QR Code Generated via (info.rules)

MetaFlows has added 2024291 to our priority alerts category, and may also add 2023218 to add an extra level of alerting for these events.


Many of the windows related scan rules have been updated, and may be treated with greater suspicion, but are not alone indicators of this malware:

2001569 – ET SCAN Behavioral Unusual Port 445 traffic Potential Scan or Infection (scan.rules)
2001579 – ET SCAN Behavioral Unusual Port 139 traffic Potential Scan or Infection (scan.rules)
2001580 – ET SCAN Behavioral Unusual Port 137 traffic Potential Scan or Infection (scan.rules)
2001581 – ET SCAN Behavioral Unusual Port 135 traffic Potential Scan or Infection (scan.rules)
2001582 – ET SCAN Behavioral Unusual Port 1434 traffic Potential Scan or Infection (scan.rules)
2001583 – ET SCAN Behavioral Unusual Port 1433 traffic Potential Scan or Infection (scan.rules)

There are likely to be more updates and more information soon as researchers have time to study the samples collected so far.
Our primary signature provider, Emerging Threats, maintains a mailing list where these issues are discussed as they unfold.

Watch your MACs

We added a feature to alert you whenever a new MAC address is seen by the system. The system learns about MAC addresses either through analyzing the DHCP protocol or finding new MAC addresses in the normal network traffic (if you are mirroring/spanning the endpoints’ MAC addresses).

It generates messages of the form:

MACwatch <IP_ADDRESS> <MAC_ADDRESS> <Flow information>

Every time the system sees a new MAC address.

IP_ADDRESS is the address using the newly discovered MAC
MAC_ADDRESS is the new MAC
<Flow Information> is the flow information we used to discover the new MAC (typically a DHCP lease, but it could also be a UDP or TCP packet if you will span the MAC addresses from the switch).

See a screen shot from our lab firewall sensor.


After the update, you will start getting messages of MAC addresses never seen before. After a while, only new MAC addresses never seen before will start showing up and you can setup a classification matching MACwatch to email yourself, block communication, or both.

The MAC addresses are available for search in the assets page under a new column called MAC. The same IP address can have multiple MACs simultaneously; and MACs can move around from IP to IP due to DHCP leasing. But, no matter what, a previously unseen MAC will generate a MACwatch message. Some devices (like printers) can go to sleep for days; so you might see some legitimate MACwatch messages for a while.

As always, let us know if you have any questions at

Happy hunting!


The MetaFlows Team

Got MAC?

We recently added the MAC addresses to the event messages. The system gets the MAC addresses in two orthogonal ways:

  • We sniff the MAC headers from the passive tap. If the MSS sees more than 5 IP addresses with the same MAC, it stops recording because it means you are mirroring the connection between the switch and the next routing hop (probably the firewall) where the MAC addresses are not available.
  • We sniff DHCP lease messages when the IP is assigned dynamically. In order to do this, you probably need to instruct the switch to specifically mirror DHCP traffic in order for the sensor to process it. The sensor expects DHCP UDP traffic using the pcap expression udp and (port 68 or port 67).

Please contact us at if you need help in setting up DHCP traffic monitoring.




Websockets Are Here

Adobe Flash is one of the original sins. It is everywhere and yet it is a huge security risk. Websockets is an HTML5 standard that, for us, provided an alternative to Adobe Flash.

For now we support both. The browser will try to use Adobe Flash first, and if it is not present or it is disabled, it will try using Websockets (which are hard-coded in your Browser). I you want to keep using Adobe Flash, you do not have to do anything; things should keep working as before.

If your sensors are configured as clients, and you do not want to use Flash anymore, just disable it and the Browser will do the rest. You will be using MetaFlows SSL certificate.

If your sensors are configured as servers, and you do not want to use Flash anymore, well, it’s a bit of work to use Websockets because current Browser implementations do not allow self-signed SSL certificates (this is probably a good thing). To use Websockets on sensors configured as servers:

  1. Add your sensors’ static IPs to the DNS (like: <sensorname1>
  2. Generate a valid SSL certificates that matches the DNS name in step 1 (cannot be self-signed). If you do not want to generate a separate certificate for each sensor, you can also buy a * certificate to share by all your sensors.
  3. Bundle the certificates with the command:

# cat my_certificate.crt my_certificate.key bundle_certificate.crt > sensorcert

  1. Copy sensorcert to /usr/local/etc/ntop/sensor-server.pem on your sensors’ hard disk.
  2. Go to and replace the static IP address of your sensors as a server with the names you setup in step 1
  3. Make sure your browser can reach the sensors on ports 3009 and 3010
  4. Restart your sensors as a server with the command:

#/nsm/etc/ restart

Adding Websockets support was a fairly extensive change in our system; so there could still be some issues. As always feel free to contact us if you have any questions or you see any problems.

Thank you for exploring the unknown with us!

The MetaFlows Team.

Constant Companions: Giving Passwords and Passphrases Thier Due

“Through 20 years of effort, we’ve successfully rained everyone to use passwords that are hard for humans to remember, but are easy for computers to guess.” Randall Monroe, XKCD

For users, passwords and passphrases are a way of life. How else can an individual not only identify themselves to access necessary services but also prove that they are who they say they are without biometrics? However, the way in which many businesses choose to think about passwords and passphrases is not only wrong, but harmful. Many financial institutions, as well as work places, require that passwords max out at a short, fixed number of characters (anything between six and twelve), include an uppercase and lowercase letter, as well as at least one digit. This is, unfortunately, not an ideal solution. In essence, any organization requiring that users make passwords under such conditions is setting their users up for failure on a multitude of levels. Not only are these passwords easier to crack than other options but they typically cannot be memorized, requiring the user(s) in question to write them down or store them elsewhere.

Data released in a recent study by Carnegie Mellon University’s CyLab indicates that traditional methods for password and passphrase creation are woefully inadequate and that a great many users have a mistaken idea of the methods in which adversaries employ in the attempt to crack them.

This study reveals that, “Participants, on average, also believed any password with numbers and symbols was a strong password, which is not always true. For example, p@ssw0rd was thought to be more secure than pAsswOrd, but the researchers’ attacker model predicted that it would take 4,000 times more guesses to crack pAsswOrd than p@ssw0rd. In modern day password-cracking tools, replacing letters with numbers or symbols is predictable.”

The question then becomes, what can the user do to avoid this situation? The engineers at MetaFlows have a very unique way of creating passwords/passphrases that are much more secure. There is a basic equation for password strength, failing that the password appears in a known dictionary, is:


Complexity being the number of possible characters the password contains

So a password using only lower case letters has a complexity of 26

A password using lower, upper, and numbers has a complexity of 62

A complex password with a length of eight:

62^8 = 218,340,105,584,896 possibilities

A simple password with a length of twelve:

26^12 = 95,428,956,661,682,176

The longer, but simpler, password in this example has a total search space 437 times greater than a standard “complex” password. This is not to say that complexity is bad, complexity helps, but length is the dominant factor in determining strength against brute force. It should be able to be memorized, so going ahead and adding a number or a weird character is fine. However, if adding that element makes it too hard to remember then consider tacking on another word instead that is easier to remember to increase the strength exponentially.

What is the difference between a password and a passphrase?

The example password meets all standard complexity requirements: lower case, upper case, number, and special character. One of our engineers decided to see how long it would take for them to crack this password. The end result is as follows:


Search Space 6.70×10^15

Single Machine traditional estimated crack time: 18.62 hours

Cracked during several hours while playing WoW and a good night’s sleep.

The experiment was repeated with a passphrase, which is a group of words strung together that act as a password. The passphrase below meets none of the standard complexity requirements as it is all lower case and contains no digits. Unlike Pa$sw0rd, it is easy to remember.


Search Space 4.33×10^39

Single Machine traditional estimated crack time: 13.76 million trillion centuries

Still not cracked long after the death of our solar system.

In most cases, adding a few words that are related to the site or process in question is helpful to remembering them but we also know that people are surprisingly good at remembering almost any silly combinations of words as a passphrase. The more unrelated the words chosen are, the less likely they will ever end up in a dictionary. Picking one nonsensical word increases the potential strength against dictionaries to a level that is realistically beyond guessable. For example, “mypasswordisnotpassword” may be obvious enough to get added to a dictionary, but “mylongitudinalpasswordisnotamonkey” is arcane.

Another method, advocated by Micha Lee at The Intercept_ is Diceware. The method for creating a Diceware password is simple and straightforward but the end results may lead to a far more secure passphrase. The Diceware method is effective because it will provide randomness that the human brain cannot. The value of using a method that involves randomization is ideal when one considers entropy. “The amount of uncertainty in a passphrase (or in an encryption key, or in any other type of information) is measured in bits of entropy. You can measure how secure your random passphrase is by how many bits of entropy it contains. Each word from the Diceware list is worth about 12.92 bits of entropy (because 212.92 is about 7,776). So if you choose seven words you’ll end up with a passphrase with about 90.5 bits of entropy (because 12.92 times seven is about 90.5).”

Once a user creates a password, one must have a clear idea of where to store it. While there are numerous password saving applications available on the web and scraps of paper abound, nothing is more secure than pure memorization. When considering password creation, always stick to something easy to memorize as well as difficult to crack. To put it plainly, storing passwords anywhere other than the human mind creates an exploitable vulnerability. This of course, includes writing them down on a sheet of paper and attempting to hide it. The popularity of password storage books and password applications is no indication as to the level of security they provide, which is limited at best.

No matter how random and entropic a password may be, it is vital that if using the same password for more than one service, that passwords used for social media accounts should in no way resemble those used for online banking and other vital activities. It cannot be stressed enough that reusing passwords, sharing passwords, recording passwords, and repeatedly recycling through a set of passwords is far from advisable.


Product Update: Reconsider Event Classifications

Recently, the engineers at MetaFlows have improved the Event Classification Menu within the MetaFlows software, allowing each user to further customize events through actions and event views. This introduces four key features to the Event Classification Menu that users will find helpful in employing the MetaFlows IDS.

Classifications Window.png

The first improvement allows users to see a comprehensive list of their classifications. Now, users can access a new classification interface that breaks the classifications down by action. There are seven action types: Highlight, Block, E-mail, Ignore, Delete, Rank, and Disabled. The Highlight function matches the records in the Real-Time, Historical, and Reports with the selected color. The Block action triggers the Soft IPS for matching records, causing connections matching the classification to be blocked. The E-mail function produces a PDF report of matching records that will be sent every ten minutes, or as frequently as possible. The Ignore action ignores events that match the classification. The Delete function removes matching records from the browser in order to free up memory. The Rank action increases the priority/rank of the records that match the classification. The Disable function allows a user to disable a classification without deleting it.


The Search functionality of the classification interface now allows users to search against a classifications’ name, category, IP address, IDS alerts, service alerts, and log message values. All a user has to do is type a value into the Search field to find classifications to match that query. The search will match against values in the classification name, category, addresses, and events field.


Once upon a time, deleting a classification was an irreversible action. Now, that can be undone. If the user deletes a classification only to realize later that they need it, they can restore the classification from the Trashed Classifications list.

Transferring classifications is now much easier. By employing the Upload Classifications feature, a user can transfer classifications in bulk between two different domains. The option is listed as the Upload Classifications button and selecting this opens the uploader. Classifications must be in JSON format and contain all the required information for the classification.

More information regarding the recent improvements in the Event Classification menu can be viewed on the MetaFlows User Manual. If using any of the four new features causes any confusion, or if there are any questions, do not hesitate to contact the MetaFlows team for assistance.