QuillHub
QuillHub
EN

Zeek Community Call — October 2025 Summary

29:154 спикера
28:59
длительность
10
главы
4
спикеры
10
Термины
Zeek developmentZeroMQCommunity contributionsOracle/TNS parsingSecurity OnionSIEM integrationTraining announcement

Спикеры

4
Спикер A33%5 реплик
Спикер B14%4 реплики
Спикер C5%1 реплика
Спикер D47%4 реплики

Основные идеи

5
  • Zeek 8.1 development is in progress with a plan to switch to ZeroMQ as the default cluster backend.
  • The project will standardize third-party integrations primarily on WebSockets or Zeek.js going forward.
  • Community contribution is essential; a new blog post outlines five easy ways to contribute for first-timers.
  • The team will launch an initiative to surface real-world Zeek usage stories ('Zeek in the wild') on Slack.
  • George demonstrated a practical custom parser for Oracle/TNS traffic to provide DB context and detections into SIEM via Security Onion.

Важные моменты

6
  • ZeroMQ will become the default cluster backend in Zeek 8.1.
  • Planned point/patch release 8.0.2 in about two weeks.
  • In-person Zeek training at the NSF Cybersecurity Summit on Oct 20; registration closes Oct 6 and the session will not be recorded.
  • New blog post: '5 Ways to Contribute to the Zeek Project for the First Time.'
  • Community experiment to share hardware setups, log workflows, and 'Zeek in the wild' stories announced for later this month.

Решения

4
  • Make ZeroMQ the default cluster backend in Zeek 8.1.
  • Standardize third-party integration approaches on WebSockets or Zeek.js going forward.
  • Schedule a patch release 8.0.2 in approximately two weeks.
  • Deliver the NSF Cybersecurity Summit Zeek training as an in-person only event and do not record it.

Задачи

4
  • Johanna to prepare a blog post and post election results to Slack regarding the LT election.
  • Interested participants to register for the in-person Zeek training at the NSF Cybersecurity Summit before registration closes.2025-10-06
  • Michelle to add the NSF Cybersecurity Summit training registration link to the YouTube video description.Michelle
  • Community members building third-party integrations (WebSockets, Broker/Python bindings, libbroker, Zeek.js) to reach out to Christian and Benjamin to discuss migration and lessons learned.

Термины

10
Zeek — Open-source network security monitoring framework for extracting network metadata and detecting suspicious activity.
ZeroMQ — High-performance asynchronous messaging library planned as the default cluster backend in Zeek 8.1.
WebSockets — Web protocol for bidirectional communication over a single TCP connection; one of the intended standard integration mechanisms.
Zeek.js — A JavaScript-based integration path for interacting with Zeek, also highlighted as a standard option.
Security Onion — Open-source platform for network security monitoring that integrates Zeek; used by George in a distributed/grid deployment.

Основное содержание

10 глав · 28:59
0:00

1. Introduction & Agenda

Спикер A0:00

Alright, welcome everybody. Thank you for tuning into our October 2025 Zeek community call. These calls are hosted by members of the Zeek leadership team, and they're open to anyone in the community to join in. We typically share project updates, training news, a couple of community reminders, and the rest is pretty flexible.

If you ever have questions or feedback, we invite you to hop on these calls and speak up. Or if there's something you'd like to see at a future meeting, we also invite you to share those thoughts here as well. We're open to feedback, and we're all ears, always.

So after our team updates today, we are going to be passing the mic to George, who is one of our community members who volunteered to share a real-world user story in the form of a lightning talk, which I myself am very excited to hear and to learn from George.

If you watched or tuned into last month's community call, you know we had Vin Stauffer give a quick talk about Zeek and post-quantum cryptography, which was super cool for us to listen to and learn from. If you did miss that talk, you can check it out on our YouTube page.

So anyway, George is gonna give a talk about something he's been doing with Zeek recently, and this is kind of an effort to highlight and share more of these real-world stories about how people are using Zeek and what they are, you know, doing with it, which is a piece of feedback that we've heard quite frequently.

And, you know, it's something that I know people are interested in getting a glimpse into how others are using Zeek, and we're hoping to kind of use this as a platform for that. So that's just a quick preview of what's to come in a little bit. But first, we have some quick updates and agenda items, and I believe Christian is going to be kicking things off today.

2:15

2. Leadership and Development Updates

Спикер B2:15

Yep, I'm subbing for Johanna today, who's on the road, and she just asked me to give a quick update, which is essentially that we have a result for the LT election in the next few days. I think she'll be probably putting together a blog post and then go into Slack regarding those. I don't have the results themselves. I think they're still in the making, but I think the election has concluded. So look out for an update.

On the development side, there isn't all that much to cover right now, basically just because we're right in the middle of our, you know, 8.1 development cycle. So that's in progress. Let me see if there's anything sort of in particular.

So essentially, yeah, we're tying up sort of the last loose ends to be able to switch to ZeroMQ as our default cluster backend in 8.1. I think that's the last thing. That's on track. Benjamin and I are very eager to talk to anyone right now who is building third-party integrations with Zeek that are using, you know, whichever of the many mechanisms are out there right now to do this.

So if it's, you know, the existing WebSockets, if it's the Python bindings for Broker, if it's libbroker itself, if it's something else, sort of, because going forward we're essentially going to standardize on either WebSockets or Zeek.js as sort of the two big ways to do this. So if you're doing something sort of adventurous there, something that you find difficult, or if you'd just like to share lessons learned for something you've built, we're all ears right now.

Would be really timely on that front. And other than that, we'll probably have a point release, a patch release 8.0.2 coming up in about two weeks, just because, you know, like the first things are sort of piling up. So stay tuned for an announcement regarding that.

And I think that's all I have today. Thanks, Michelle.

4:10

3. Training Announcement (NSF Cybersecurity Summit)

Спикер A4:10

Thanks, Christian. Fatima, do you have anything to share regarding training?

Спикер C4:18

Yes. So just a quick reminder, we are pretty excited to have our in-person Zeek training at the NSF Cybersecurity Summit this year. That's gonna happen in 20 days. Registrations are still open, but they are not going to be open for long, so they are gonna close on the 6th of October.

So, people who are interested in coming in person, meeting us all, and attending the training, please do so. Register before the registration closes. The training will be the first day of the NSF Cybersecurity Summit, which is going to be the 20th of October.

So, it's a full-day training, and again, a reminder, it will not be recorded, and it will be delivered in just in-person format. So, whoever out there is willing to come attend, say hi to us, we are more than happy to kind of like host this training, and we are always as excited as we can because this is one of our chances to meet and greet the community and to know what's going on in the users' and Zeek community people's lives.

So, don't miss this opportunity, and if you have any questions, please reach out to us. NSF Cybersecurity Summit, I think we have shared that link before in some of the Zeek channels, but if you have missed it and if you don't know where to find or go for registration, please poke one of us, and we would be more than happy to help you guide through the registration.

So, that's an update from my side. Thanks, Michelle.

5:48

4. Community Engagement and Contribution Initiatives

Спикер A5:48

Awesome, thanks. I'll be sure to put a link to the registration in the YouTube description, so for the people who are watching the recording, they can just scroll on down and they'll have all that information right there.

All right, so I have just some quick updates this month before we pass it over to George for our lightning talk. Both of my community updates are related to engagement and contribution. You know, I've spent some time the last month really beginning to dig into the feedback from the community survey that we did over the summer. And I'm starting to really focus on some specific recurring pain points or obstacles that I've been noticing in the feedback of the community experience.

Contribution is a big one that I'm seeing people talk about a lot. And, you know, I do want to mention, as an open-source project, Zeek really does depend on the community for support and for growth in a number of ways. I really want to make sure that everyone in our community knows that. So when we asked about contribution in the survey, we saw about a 50/50 split between people saying yes, they've contributed before, in some way, versus people who have not done so.

So I published a blog post this morning on the Zeek website. It's called "5 Ways to Contribute to the Zeek Project for the First Time." Very direct, very specific. And it's a really quick read for that 50-51% of the community who say they haven't contributed yet. It addresses some of those specific barriers that people mentioned to contribution, whether that's, you know, not enough time or bandwidth, not feeling qualified or experienced enough, or not knowing where to start.

I kind of take each of those barriers and give very specific entry points to being a Zeek contributor and playing an active role in the community, whether you are a builder or developer, or just a Zeek user, if you're new, or if you've been around for a while. If you are curious about getting a bit more involved, please, you know, give that blog post a quick look. It's a quick scan. I do think that there's something for everyone here.

And, you know, if you do take a look and you want to get involved, but the 5 ways that I suggest aren't it, definitely send me a Slack message and let me know, and we can have a conversation about what might interest you. But over the next several months, I really am hoping to connect with more community members, get more people involved, and bring some more engagement and kind of energy into this space. I think there's a lot of really awesome engagement and energy happening right now, and I think we all just want more of it because it's really great what we have so far.

So that was update number 1. Check out that blog post. Update number 2 is adjacent to this. So, you know, I mentioned earlier in the call with, you know, George joining us for a quick lightning talk about how he's using Zeek.

So something the leadership team has been talking about recently is how we might provide more opportunities for people to share and discuss their experiences with Zeek, you know, really as a way for people to learn from one another and kind of what are their Zeek in the wild stories, what are they— what successes are they having? What mistakes are they making? And everything in between.

So later this month, I'm going to launch an initiative that will give people the chance to talk about things like hardware setups, how you are working with logs, or even just kind of surprise aha moments. I think there's going to be a lot of different opportunities for people to say, oh, that actually does apply to me, um, let me join the conversation and share.

So the goal really is to, you know, spark discussion in the community and make it very easy to learn from other people's experiences. So keep an eye out later this month; we'll be sharing information on the Zeek Slack workspace. That's kind of where we'll be sharing all of the information and the details that you'll need about this experiment that I'm excited to try out.

All right. So blog post, Slack announcement coming soon, and that is all I had on the community front. So thank you.

11:34

5. Lightning Talk: Speaker Introduction

Спикер A11:34

All right, George, I am going to pass the mic to you. The floor is yours for the rest of the call, and we're excited to kind of hear about what you've been doing with Zeek recently.

Спикер D11:49

Thank you, Michelle. Thank you, Christian. I'm very glad to participate in this community call. As I mentioned before, it's going to be my first contribution in any open source community, which is— I'm glad doing it because it's going to be helping a lot of people and a lot of guys in that community to improve themselves and improve the product that we're working with, which is going to be Zeek.

I'm going to start with a small presentation that I have prepared. So I'm going to do a small presentation. I'm a cybersecurity consultant with more than 20 years in IT security. It's been more than 10 years since I switched to cybersecurity. Before, I was a Cisco instructor and IBM System X engineer, and it's been 25 years teaching Cisco and teaching networking at the university. So I switched to cybersecurity within 10 years. Now I'm specialized in cloud security, SIEM, SOAR, and automation. I gained some experience in Security Onion, ExtraHop, Sentinel, FortiGate, and those only.

I'm mostly passionate about building custom parsers. It's coming from an old-school programming background since I started programming with COBOL for mainframe in 1998, so I'm still passionate about the programming and the coding that I've been doing. It's a hobby for me.

What led me to Zeek? The first thing, we needed to install a redundant NDR to some kind of a product that was installed in the company. When we started deploying the Security Onion, we noticed that Security Onion doesn't provide any visibility for TNS Oracle traffic. And that was very critical for the security operation of that company. We faced that problem, and the full project was halted and stopped until we solved that problem.

13:51

6. Monitoring Oracle/TNS Traffic with Zeek

Спикер D13:51

So what I started to think of doing, I went into Zeek and I tried to create some kind of a parser to intercept any SQL traffic. By intercepting that SQL traffic, I tried to extract from it the connection that was done from the source IP, destination IP, and source port, destination port, and to extract the SQL statement issued inside that connection or the commands. I tried to put any SQL statements that exist inside Oracle that can be added because it's a static list inside the script. I'm looking at all SELECT, INSERT, CREATE, GRANT, REVOKE, and anything that's going to be related to a security issue inside that traffic.

Then I went furthermore to get all the errors that are going to be raised during that connection. So we're going to have a TNS error or we're going to have an ORA error. It's going to be related to a system error or a command executed. In that way, we can monitor everything or all commands done through that SQL statement and through those connections to raise the alerts that we need.

Now, I got that by creating several logs and using Zeek, and I also tried to create a notice based on critical Oracle errors that in my case were considered as critical. It's in a static list inside the ZK script that can be tuned, added, or you can remove. So I'm producing three logs. I'm producing the tns.log that's going to contain the connection itself. It means it's going to contain the source IP, destination IP, ports used, and the user used in that connection.

What I'm doing to remove the applications, because inside the connection I'm going to get multiple logs, and inside that connection I'm going to get also multiple SQL queries or multiple errors related to the same connection. I'm extracting those logs as connection, as username, as SQL, and error. I'm putting it in a table. That table is going to be created once the connection is opened using the event tcp_connect of Zeek.

During that same connection, whenever I'm going to face the same error or the same SQL or the same user of the connection, I'm not going to take it into consideration. I'm going to drop it. Once the connection is going to be closed inside the event connection state remove, I'm destroying that table. In that way, I can now correlate the SQL statement with the user, and I can correlate even the error with the users that created that connection.

I faced one problem and I couldn't do it well. It was the correlation of the SQL statement with the error produced because I might have two SQL statements in the same connection that's going to be providing perhaps one error or two errors. So I could not correlate it. But it's going to be easy as manual interception. I can get the connection ID with the user and can go to the errors and can get the connection ID also and the error and the SQL statement and can do that correlation.

In that way, I'm going to be able to detect any suspicious or unauthorized DB activity. I can get all the detections I'm going to be creating. It's going to be created on our SIEM. Our SIEM, we're using Microsoft Sentinel SIEM. So any analytic rules or alerts are going to be raised in that Sentinel.

17:27

7. Logs, SIEM Integration, and Implementation Details

Спикер D17:27

And I'm raising some kind of alerting inside the notice.log of Zeek. It's based on specific Aura or TNS errors. Now, what I'm going to be having inside my tns.log, I'm going to be having, it's an example here. I'm putting that I have the connection, I have the source IP, I have, sorry, I have the one, I have the source IP with the port used, destination IP, the SQL server, the port, and the user.

And inside the TNS SQL log, I'm putting all the SQL statements executed. So it's going to be the select, an update, a drop, alter, whatever is going to be executed is going to be here. And I created the third log. It's going to be TNS error.log. It's going to have the same, same information as source IP and connection, the port, the user, and what were the errors in the Aura and the description of that error.

Now, the problems that I faced also, it was sending those logs to my SIEM. Since I'm using Security Onion and security— I'm using Security Onion in a distributed mode on a grid, I faced that problem of sending those logs to my SIEM. I could have used the Zeek2CEF Python scripts. I can use the ZeekCEF plugin to send it directly to our SIEM.

But I was afraid of altering or changing the configuration in Security Onion. So that's why I looked for an on-top solution to leave Security Onion as is, not to touch any configuration inside it, and create a Linux script to read the notice.log as tailing to detect any new detection I'm going to receive.

And I'm going to parse it because it's going to be JSON inside the notice.log since it's Security Onion. Zeek is going to be part of Zeek Tanya. So I'm going to be parsing it, I'm going to be preparing the CEF message, I'm going to send it to the CEF syslog.

And that's, that's the script I'm going to create as a service inside my Linux to run continuously, and I'm going to get all the logs and all the errors that are going to be raised from that system script to double-check it.

19:46

8. Project outcomes: Oracle monitoring with Zeek and Security Onion

Спикер D19:46

What, what, what value I gained from that Zeek script? First of all, I improved the monitoring of Oracle-based applications since I'm working in a company that it's critical for them to monitor any Oracle-based application, any Oracle traffic, because it's going to be the first part of an exfiltration.

So any access to any critical information, to any sensible, sensitive information into the database, it should be monitored from scratch. So I should see which user did the exfiltration, or which user did the select from the database, or he did, he tried to do some corruption of the tables by issuing commands like alter or dropping multiple tables from the same user.

So all of these are use cases that can be created on the SIEM. I did not create them inside the Zeek script to raise alerts in notice because it's gonna be really, it's gonna be dependent on the SIEM. It's gonna be dependent on the analytic rules and the use cases gonna be created on my SIEM.

And it's gonna be, it's gonna happen using a correlation with other events and other detections. That's why I did not create any use cases inside Zeek. Only the, I read some alerts for critical errors that's gonna happen.

So I enhanced the correlation in SIEM by adding the DB context. I helped implement Security Onion as an enterprise-grade NDR because enterprise and enterprise-grade NDRs are gonna— in their marketing, they're gonna say that we're gonna be monitoring DNS products, we're gonna be monitoring direct traffic.

It's kind of a marketing issue. So here I took an open-source solution, Security Onion, and I raised it to another grade to be able to monitor the Oracle. So just a small thing for the community. My custom parsers made that project of Security Onion to be a redundant NDR, make it happen, make it a reality.

Now it's really a redundant NDR. And we're catching some more detections, more than the production NDR and the other NDRs that exist. I'm gonna try to expand protocol coverage by getting some other, by creating some other scripts and doing some other parsers.

What I'm gonna try to do also, I'm gonna try to add to this parser some kind of behavior analytics. So I'm gonna try to do some machine learning to get which server or which user is gonna be doing each day some select statements on that DB.

So I'm not gonna raise any detection for that. And for any new connection, for any new or none seen before connection or a select statement, I'm gonna raise an alert for that. I started small, so it was, it was like an idea. Can we, can we install Security Onion? My answer was yes.

As an IT person, we never say no for anything. We always say yes. Uh, it took me a lot of tests because, uh, I'm working inside Security Onion. I tried my— I started my scripts on Zeek 8.0. I moved it to Security Onion. I faced problems that Security Onion is going to be working with Zig 7.0.10 most recently with the newest update.

So, I had to start from scratch. I did a lot of tests, and now I'm sharing it with the community so everyone can take advantage of that. Thank you.

23:54

9. Community Q&A: Parser development and technical feedback

Спикер A23:54

Alright. So, yeah, if anyone has thoughts or questions, or, you know, just kind of wants to recognize George for, you know, volunteering to do this, feel free to chime in.

Спикер B24:09

Yeah, thank you, George. I wish we had more people who offer, you know, that amount of information about the Zeek work that they do. So, like, maybe you're, you know, a trailblazer there. So that's really cool.

I'm curious. So, okay, so I have questions about the protocol itself and then how you decided and figured out how to go about writing a parser for it. So the first one is probably pretty easy, but that protocol is proprietary, I think, right? Like, so how did you figure out how it works?

Спикер D24:39

Yeah, 100%, it's proprietary; it's Oracle. So I started using, doing some Wireshark PCAPs, just to understand how it was and the structure of it. Sometimes you have that traffic is going to be encrypted. I'm going to be working on how to decrypt it and how to get the session keys from the Oracle database to decrypt that traffic.

So it's—I know it's proprietary, but the documentation of Oracle is mentioned obviously and clearly how that packets and what the traffic is. And it was on trial and error, Christian. So it took me a lot of time just to understand how it was built and what fields I have to extract and the amount of fields I have to extract.

And also, I faced some problems in the raw data itself. I need to do some cleanup to remove the non-readable characters and to remove the null bytes and all of that. So yeah, it took me a little bit of time to do it, but it was good.

Спикер B25:40

Yeah, it sounds great. That's exactly what, you know, we would do as well if we just had to sort of figure it out.

And then, so in order to write a parser, there are sort of several, you know, established ways in Zeek of doing this, and your input there would be very welcome.

So did you find sort of available options there? Did you maybe consider writing one in Spicy, which is sort of our sort of, you know, stock, you know, in-house parser?

Yeah, I did. Parser technology, or did you essentially just sort of like use the script layer because it was the most accessible thing?

Спикер D26:20

No, not really. I used the script layer directly, so I did not use the spaCy because I'm already, like I mentioned, I already have some experience in coding and in the old school coding.

So it was not related to any debugger or any parser for it.

So that's why it was on trial and error. I was creating my script, running my script, seeing what was the error, fixing the error, and running the script from scratch.

Спикер B26:53

Yeah, yeah, so the danger of doing it fully in the script layer the way you did it is that you're effectively almost bypassing the analyzer framework. That's a total, like, low-level, you know, thing in there that we haven't documented as much as we probably should.

But, so if you have other tooling out there that sort of, you know, understands what analyzers are present in Zeek and so forth, then yours is currently not sort of visible there because you're essentially, you know, not hooked in there as you would be if you were using, let's say, spaCy.

So if you're interested, we can send you some feedback on that level and help you get it sort of like in that sort of position. But, you know, because it's also like, it's clear, like, like, like, like using spaCy would have meant sort of a different learning curve. You would have had to sort of like, like figure that one out as well.

But there's much scope, I think, and it clearly, you know, super relevant protocol. So that, that's, that's excellent. So thank you. Thank you. I'm glad you figured it out to the point where it got you the logs that you needed.

So that sounds like a great win. That's cool. So yeah, thanks again. That's all I had. Thank you.

Спикер D27:52

Thank you, Christian.

Спикер A27:56

Awesome. Thanks, Christian. And thank you, George, again. That was awesome.

28:01

10. Meeting wrap-up and next steps

Спикер A28:01

All right, so we're going to wrap things up. If you're watching this on YouTube, check out the description. I'll, you know, make sure that there are plenty of links for all of the things that we mentioned in this month's meeting.

The next community call is going to be on November 5th, 10 AM Pacific time. As always, we kind of share some reminders on Slack and on our social media channels. So keep an eye out for that.

And if you are doing something with Zeek and wouldn't mind spending 10-15 minutes just talking about it, we're always looking for community members to jump in and just kind of share with us what they've been doing.

So thank you again, George, and thank you everyone for tuning in. Thank you. We'll see you next month.