Yevgeniy Brikman
@brikis98.bsky.social
Co-founder: https://gruntwork.io. Author: https://fundamentals-of-devops.com, https://terraformupandrunning.com, https://hello-startup.net. Previously: LNKD, TRIP.
I just finished "The Anthropocene Reviewed" by John Green. It's a collection of essays on a variety of topics from the modern era, including the QWERTY keyboard, Canadian geese, Super Mario Kart, teddy bears, and Haley’s Comet.
A few thoughts:
A few thoughts:
November 7, 2025 at 3:15 PM
I just finished "The Anthropocene Reviewed" by John Green. It's a collection of essays on a variety of topics from the modern era, including the QWERTY keyboard, Canadian geese, Super Mario Kart, teddy bears, and Haley’s Comet.
A few thoughts:
A few thoughts:
In 2022, I got a desk treadmill, and it turned out to be one of the best investments I’ve made in my health in years. It improved my fitness, productivity, and likely added years to my life. I’ve gotten so many questions about it that I decided to write a blog post:
November 6, 2025 at 1:50 PM
In 2022, I got a desk treadmill, and it turned out to be one of the best investments I’ve made in my health in years. It improved my fitness, productivity, and likely added years to my life. I’ve gotten so many questions about it that I decided to write a blog post:
My first task at TripAdvisor was to add a new sort option to the hotels page. Easy! I got it done & deployed in my 1st week. I then had my first 1:1 with my manager. He opened the hotels page for Paris, picked my new sort option, and waited. And waited. And waited...
October 21, 2025 at 6:48 PM
My first task at TripAdvisor was to add a new sort option to the hotels page. Easy! I got it done & deployed in my 1st week. I then had my first 1:1 with my manager. He opened the hotels page for Paris, picked my new sort option, and waited. And waited. And waited...
I've come across many companies where:
🔥 They're constantly putting out fires
🚨 Nothing is reproducible or documented
⚠️ Even minor changes feel risky
😰 The predominant emotions are fear & stress
This is "DevOps Bankruptcy." We wrote a guide on how to get yourself out of it:
🔥 They're constantly putting out fires
🚨 Nothing is reproducible or documented
⚠️ Even minor changes feel risky
😰 The predominant emotions are fear & stress
This is "DevOps Bankruptcy." We wrote a guide on how to get yourself out of it:
October 8, 2025 at 3:11 PM
I've come across many companies where:
🔥 They're constantly putting out fires
🚨 Nothing is reproducible or documented
⚠️ Even minor changes feel risky
😰 The predominant emotions are fear & stress
This is "DevOps Bankruptcy." We wrote a guide on how to get yourself out of it:
🔥 They're constantly putting out fires
🚨 Nothing is reproducible or documented
⚠️ Even minor changes feel risky
😰 The predominant emotions are fear & stress
This is "DevOps Bankruptcy." We wrote a guide on how to get yourself out of it:
Here's a list of the best resources related to Chapter 9, "How to Store Data" of "Fundamentals of DevOps and Software Delivery." Learn about relational DBs, NoSQL, NewSQL, queues, streams, object stores, CDNs, backup, recovery, and more:
October 6, 2025 at 1:44 PM
Here's a list of the best resources related to Chapter 9, "How to Store Data" of "Fundamentals of DevOps and Software Delivery." Learn about relational DBs, NoSQL, NewSQL, queues, streams, object stores, CDNs, backup, recovery, and more:
I just completed a fun project: I put all my reviews & notes from the books I've read on my blog!
I read ~100 books / year, and I've been tracking my reading for more than a decade, so this represents my ratings and notes for over 1,000 books.
Links are in the following tweets:
I read ~100 books / year, and I've been tracking my reading for more than a decade, so this represents my ratings and notes for over 1,000 books.
Links are in the following tweets:
October 2, 2025 at 1:14 PM
I just completed a fun project: I put all my reviews & notes from the books I've read on my blog!
I read ~100 books / year, and I've been tracking my reading for more than a decade, so this represents my ratings and notes for over 1,000 books.
Links are in the following tweets:
I read ~100 books / year, and I've been tracking my reading for more than a decade, so this represents my ratings and notes for over 1,000 books.
Links are in the following tweets:
Here's a list of the best resources related to Chapter 8, "How to Secure Communication and Storage" of "Fundamentals of DevOps and Software Delivery." Learn about cryptography, encryption in transit and at rest, hashing, secrets management, TLS certificates, AES, and more:
September 1, 2025 at 11:59 AM
Here's a list of the best resources related to Chapter 8, "How to Secure Communication and Storage" of "Fundamentals of DevOps and Software Delivery." Learn about cryptography, encryption in transit and at rest, hashing, secrets management, TLS certificates, AES, and more:
CI/CD servers can be a huge security risk:
1. They have powerful credentials
2. Every dev in your company has access
3. They are used to run arbitrary code
That's a terrifying combination from a security perspective 😳😱
Here are 8 steps for secure CI/CD:
1. They have powerful credentials
2. Every dev in your company has access
3. They are used to run arbitrary code
That's a terrifying combination from a security perspective 😳😱
Here are 8 steps for secure CI/CD:
August 18, 2025 at 4:40 PM
CI/CD servers can be a huge security risk:
1. They have powerful credentials
2. Every dev in your company has access
3. They are used to run arbitrary code
That's a terrifying combination from a security perspective 😳😱
Here are 8 steps for secure CI/CD:
1. They have powerful credentials
2. Every dev in your company has access
3. They are used to run arbitrary code
That's a terrifying combination from a security perspective 😳😱
Here are 8 steps for secure CI/CD:
Here's a list of the best resources related to Chapter 7, "How to set up Networking" of "Fundamentals of DevOps and Software Delivery." Learn about DNS, VPCs, VPN, SSH, service discovery, and service meshes:
August 14, 2025 at 11:41 AM
Here's a list of the best resources related to Chapter 7, "How to set up Networking" of "Fundamentals of DevOps and Software Delivery." Learn about DNS, VPCs, VPN, SSH, service discovery, and service meshes:
Feature toggle deployment:
1. Start with v1 of your app.
2. Deploy v2 using a core strategy, but any new features in v2 are wrapped in a feature toggle, and off by default, so users see no difference from v1.
3. Release the new features in v2 by flipping the feature toggle on.
1. Start with v1 of your app.
2. Deploy v2 using a core strategy, but any new features in v2 are wrapped in a feature toggle, and off by default, so users see no difference from v1.
3. Release the new features in v2 by flipping the feature toggle on.
August 7, 2025 at 4:09 PM
Feature toggle deployment:
1. Start with v1 of your app.
2. Deploy v2 using a core strategy, but any new features in v2 are wrapped in a feature toggle, and off by default, so users see no difference from v1.
3. Release the new features in v2 by flipping the feature toggle on.
1. Start with v1 of your app.
2. Deploy v2 using a core strategy, but any new features in v2 are wrapped in a feature toggle, and off by default, so users see no difference from v1.
3. Release the new features in v2 by flipping the feature toggle on.
Promotion deployment:
1. Start with v1 in all envs (dev, stage, prod).
2. Deploy v2 to dev using a core strategy.
3. If everything works well in dev, deploy ("promote") v2 to stage.
4. If everything works well in stage, promote v2 to prod.
1. Start with v1 in all envs (dev, stage, prod).
2. Deploy v2 to dev using a core strategy.
3. If everything works well in dev, deploy ("promote") v2 to stage.
4. If everything works well in stage, promote v2 to prod.
August 7, 2025 at 4:09 PM
Promotion deployment:
1. Start with v1 in all envs (dev, stage, prod).
2. Deploy v2 to dev using a core strategy.
3. If everything works well in dev, deploy ("promote") v2 to stage.
4. If everything works well in stage, promote v2 to prod.
1. Start with v1 in all envs (dev, stage, prod).
2. Deploy v2 to dev using a core strategy.
3. If everything works well in dev, deploy ("promote") v2 to stage.
4. If everything works well in stage, promote v2 to prod.
Canary deployment:
1. Start with v1 of your app.
2. Deploy 1 replica of v2 (the canary) and compare it to a v1 replica (the control). If you find undesired differences between the canary & control (e.g., higher error rates), roll back.
3. Fully roll out v2 using a core strategy.
1. Start with v1 of your app.
2. Deploy 1 replica of v2 (the canary) and compare it to a v1 replica (the control). If you find undesired differences between the canary & control (e.g., higher error rates), roll back.
3. Fully roll out v2 using a core strategy.
August 7, 2025 at 4:09 PM
Canary deployment:
1. Start with v1 of your app.
2. Deploy 1 replica of v2 (the canary) and compare it to a v1 replica (the control). If you find undesired differences between the canary & control (e.g., higher error rates), roll back.
3. Fully roll out v2 using a core strategy.
1. Start with v1 of your app.
2. Deploy 1 replica of v2 (the canary) and compare it to a v1 replica (the control). If you find undesired differences between the canary & control (e.g., higher error rates), roll back.
3. Fully roll out v2 using a core strategy.
How do these core deployment strategies compare? Here's a table that shows their strengths and weaknesses across several dimensions:
August 7, 2025 at 4:09 PM
How do these core deployment strategies compare? Here's a table that shows their strengths and weaknesses across several dimensions:
Blue-green deployment:
1. Start with v1 (blue) of your app.
2. Deploy v2 of your app (green) onto new servers.
3. When all the v2 replicas are passing health checks, do an instantaneous switchover, moving all traffic from v1 (blue) to v2 (green). Then, undeploy v1.
1. Start with v1 (blue) of your app.
2. Deploy v2 of your app (green) onto new servers.
3. When all the v2 replicas are passing health checks, do an instantaneous switchover, moving all traffic from v1 (blue) to v2 (green). Then, undeploy v1.
August 7, 2025 at 4:09 PM
Blue-green deployment:
1. Start with v1 (blue) of your app.
2. Deploy v2 of your app (green) onto new servers.
3. When all the v2 replicas are passing health checks, do an instantaneous switchover, moving all traffic from v1 (blue) to v2 (green). Then, undeploy v1.
1. Start with v1 (blue) of your app.
2. Deploy v2 of your app (green) onto new servers.
3. When all the v2 replicas are passing health checks, do an instantaneous switchover, moving all traffic from v1 (blue) to v2 (green). Then, undeploy v1.
Rolling deployment with replacement:
1. Start with v1 of your app.
2. Replace one v1 replica with a v2 replica. This strategy is often used with stateful systems, so you also move the data from v1 to v2.
3. Repeat the process in step 2 until all v1 replicas are replaced with v2.
1. Start with v1 of your app.
2. Replace one v1 replica with a v2 replica. This strategy is often used with stateful systems, so you also move the data from v1 to v2.
3. Repeat the process in step 2 until all v1 replicas are replaced with v2.
August 7, 2025 at 4:09 PM
Rolling deployment with replacement:
1. Start with v1 of your app.
2. Replace one v1 replica with a v2 replica. This strategy is often used with stateful systems, so you also move the data from v1 to v2.
3. Repeat the process in step 2 until all v1 replicas are replaced with v2.
1. Start with v1 of your app.
2. Replace one v1 replica with a v2 replica. This strategy is often used with stateful systems, so you also move the data from v1 to v2.
3. Repeat the process in step 2 until all v1 replicas are replaced with v2.
Rolling deployment without replacement:
1. Start with v1 of your app.
2. Deploy v2 onto new servers. As the v2 apps pass health checks, the load balancer sends traffic to both v1 and v2 nodes.
3. Undeploy the v1 apps, until you end up with just v2 running.
1. Start with v1 of your app.
2. Deploy v2 onto new servers. As the v2 apps pass health checks, the load balancer sends traffic to both v1 and v2 nodes.
3. Undeploy the v1 apps, until you end up with just v2 running.
August 7, 2025 at 4:09 PM
Rolling deployment without replacement:
1. Start with v1 of your app.
2. Deploy v2 onto new servers. As the v2 apps pass health checks, the load balancer sends traffic to both v1 and v2 nodes.
3. Undeploy the v1 apps, until you end up with just v2 running.
1. Start with v1 of your app.
2. Deploy v2 onto new servers. As the v2 apps pass health checks, the load balancer sends traffic to both v1 and v2 nodes.
3. Undeploy the v1 apps, until you end up with just v2 running.
Downtime deployment:
1. Start with v1 of your app.
2. Take the v1 nodes down and update to v2. While the update is happening, your users get an outage.
3. Once v2 is running everywhere, your users are able to use the app again.
1. Start with v1 of your app.
2. Take the v1 nodes down and update to v2. While the update is happening, your users get an outage.
3. Once v2 is running everywhere, your users are able to use the app again.
August 7, 2025 at 4:09 PM
Downtime deployment:
1. Start with v1 of your app.
2. Take the v1 nodes down and update to v2. While the update is happening, your users get an outage.
3. Once v2 is running everywhere, your users are able to use the app again.
1. Start with v1 of your app.
2. Take the v1 nodes down and update to v2. While the update is happening, your users get an outage.
3. Once v2 is running everywhere, your users are able to use the app again.
We’re exploring a new add-on for Terragrunt that adds enterprise features, but before building it, we'd love your feedback.
See the features: postimg.cc/5jL74Nqt
Share your feedback: forms.gle/Vy8vi16WPtTC...
10 randomly selected participants will get a $20 gift card.
See the features: postimg.cc/5jL74Nqt
Share your feedback: forms.gle/Vy8vi16WPtTC...
10 randomly selected participants will get a $20 gift card.
August 1, 2025 at 9:30 PM
We’re exploring a new add-on for Terragrunt that adds enterprise features, but before building it, we'd love your feedback.
See the features: postimg.cc/5jL74Nqt
Share your feedback: forms.gle/Vy8vi16WPtTC...
10 randomly selected participants will get a $20 gift card.
See the features: postimg.cc/5jL74Nqt
Share your feedback: forms.gle/Vy8vi16WPtTC...
10 randomly selected participants will get a $20 gift card.
Here's a list of the best resources related to Chapter 6, "How to Work with Multiple Teams and Environments" of "Fundamentals of DevOps and Software Delivery." Learn about setting up environments, splitting your codebase into libraries, and monoliths vs microservices.
July 31, 2025 at 5:31 PM
Here's a list of the best resources related to Chapter 6, "How to Work with Multiple Teams and Environments" of "Fundamentals of DevOps and Software Delivery." Learn about setting up environments, splitting your codebase into libraries, and monoliths vs microservices.
Here's a list of the best resources related to Chapter 5, "How to set up CI and CD" of "Fundamentals of DevOps and Software Delivery." Learn about GitOps, trunk-based development, feature flags, zero-downtime deployments, and more:
July 14, 2025 at 7:27 PM
Here's a list of the best resources related to Chapter 5, "How to set up CI and CD" of "Fundamentals of DevOps and Software Delivery." Learn about GitOps, trunk-based development, feature flags, zero-downtime deployments, and more:
Every drug becomes toxic at a high enough dose, so you always use the "minimum effective dose": just enough to get the benefits, and no more.
The same is true with DevOps: every tool & technique has a cost, so you should use the "minimum effective dose of DevOps," too.
The same is true with DevOps: every tool & technique has a cost, so you should use the "minimum effective dose of DevOps," too.
July 7, 2025 at 4:13 PM
Every drug becomes toxic at a high enough dose, so you always use the "minimum effective dose": just enough to get the benefits, and no more.
The same is true with DevOps: every tool & technique has a cost, so you should use the "minimum effective dose of DevOps," too.
The same is true with DevOps: every tool & technique has a cost, so you should use the "minimum effective dose of DevOps," too.
Here's a list of the best resources related to Chapter 4, "How to Version, Build, and Test Your Code" of "Fundamentals of DevOps and Software Delivery." Learn version control (Git, GitHub), build systems (Gradle, Bazel), & automated tests (unit & integration tests):
July 2, 2025 at 1:30 PM
Here's a list of the best resources related to Chapter 4, "How to Version, Build, and Test Your Code" of "Fundamentals of DevOps and Software Delivery." Learn version control (Git, GitHub), build systems (Gradle, Bazel), & automated tests (unit & integration tests):
Respond with a caption for this image.
I’ll start: I’m excited to introduce my new book, “Fundamentals of CatOps and SoftFur Delivery.”
I’ll start: I’m excited to introduce my new book, “Fundamentals of CatOps and SoftFur Delivery.”
June 27, 2025 at 1:17 PM
Respond with a caption for this image.
I’ll start: I’m excited to introduce my new book, “Fundamentals of CatOps and SoftFur Delivery.”
I’ll start: I’m excited to introduce my new book, “Fundamentals of CatOps and SoftFur Delivery.”
Step 9: To handle a global user base, you set up multiple data centers around the world, and you connect those data centers over the network (VPN). To boost developer productivity and standardize coding practices, you set up an internal developer platform (IDP).
June 23, 2025 at 4:25 PM
Step 9: To handle a global user base, you set up multiple data centers around the world, and you connect those data centers over the network (VPN). To boost developer productivity and standardize coding practices, you set up an internal developer platform (IDP).
Step 8: To process all of your company's data, you set up analytics tools (e.g., data warehouse). To handle even more microservice communication, you turn to event-streaming and event-driven architectures. To make deployments more reliable at scale, you use feature toggles.
June 23, 2025 at 4:25 PM
Step 8: To process all of your company's data, you set up analytics tools (e.g., data warehouse). To handle even more microservice communication, you turn to event-streaming and event-driven architectures. To make deployments more reliable at scale, you use feature toggles.