9 ms·

Thundra: AWS Lambda Observability for Java, Go and Node.js

74 points by CSDude 9y ago

_Marak_ 9y ago
This looks promising! We tried using a similar service "IOPipe" last year and really had trouble with their offering. https://www.iopipe.com/ https://www.iopipe.com/ It would be nice to see more companies providing this kind of observability services ( especially for multiple cloud providers )
- sozal 9y ago
  Hi, Thundra has more advanced features such as - Automated instrumentation for tracing (no code change is required) - Aggregating metrics and logs with traces - Async monitoring over CloudWatch logs with almost zero overhead - Distributed tracing support by AWS X-Ray integration - and others ... Check www.thundra.io out for more
- dankohn1 9y ago
  CNCF is current tracking 8 projects/products in the serverless tools space, including Thundra and IOPipe: https://landscape.cncf.io/grouping=landscape&landscape=tools https://landscape.cncf.io/grouping=landscape&landscape=tools
bthornbury 9y ago
As an interested user, what is the difference here between traditional instrumentation like Application Insights? I understand this is intended for AWS Lambda, but what is it doing that makes it more suited?
- sozal 9y ago
  Hi, Instrumentation (manual and automated) for tracing is just one of the features of Thundra. Thundra provides all the three pillars of observability in one tool. It can aggreagate metrics and logs. Thundra is also integrated with AWS X-Ray for supporting distributed tracing. Additionally, Thundra can monitor your Lambda functions with almost zero overhead by its async monitoring functionality over CloudWatch logs.
jedberg 9y ago
I tried creating an account but it just says: "Error: server_error. Check the console for further details." :( This looks really interesting! Also, the title here says "Python", but I don't see Python in the docs. I assume it is similar to Go, where you just import a library?
- sozal 9y ago
  Hi, Can you try by enabling 3rd party cookies or whitelist Auth0's domain in your browser?
  - jedberg 9y ago
    I switched browsers from Safari to Chrome and it worked. Any info on the Python question? Thanks!
    - sozal 9y ago
      Unfortunately Python is not supported yet but it is in progress. It will be available in a couple of weeks. Thanks for your interest to Thundra
reconbot 9y ago
I've built a cloudwatch metrics stack that reads out of lambda's logs. There's no other way to get data out of your lambda for "free". XRAY is an option but not everything you'd want.
- jelder 9y ago
  I went through the trouble of setting up X-Ray for three of our Serverless projects (~50 Lambda functions) and can't say that we've gotten much value from it. The UI is... lacking.
  - emrahsamdan 9y ago
    Thanks for giving a try! We are here to continuously improve our product with your feedbacks. What specifically can you recommend to improve our UI?
thinkingkong 9y ago
This is cool! Strategically are you worried about being built on top of an AWS specific solution? The only reason I ask is its pretty inevitable that AWS will be providing instrumentation / apm capabilities native to their offering.
- sozal 9y ago
  Hi, Thundra is not designed just for AWS Lambda. Even though currently we are supporting it on AWS Lambda platform, it will be available for other environments such as containers soon.
adevine 9y ago
I'm not quite sure I understand what the log support entails. Also, it says for Node.js that this is coming soon. Any idea of the ETA on this one?
- berkay 9y ago
  You can select a method and see the logs generated for that method. Idea is to have all the information - request parameters, return values, logs, metrics all in one place to make it easier to troubleshoot problems.
biztos 9y ago
This would be super timely for me if it supported Python, but even without that it's interesting. Coming from the world of servers (virtual or physical) I find the logging latency and the lack of "observability" in general quite maddening in Lambda. Two things I wonder about this: 1) The "no code changes" part implies it's adding another layer of code somewhere, which means another thing to trust and probably not be able to audit, which is not automatically OK outside the "move fast break things" world. Am I misreading it? 2) How can you add metrics with "no overhead?" Surely there must be some overhead... and then it would be nice to be able to measure how much. For me personally, #1 is a real issue and #2 is just hypothetical, but I would expect it to be the other way around for a fast-growing startup.
- adjohn 9y ago
  If you're in a rush, IOpipe[1] already has Python support offering profiling, tracing, and other debugging and observability tools for AWS Lambda. Co-founder of IOpipe here, feel free to ask anything. [1] - https://www.iopipe.com https://www.iopipe.com
- emrahsamdan 9y ago
  You are right about lack of observability tools for AWS Lambda. Thundra was born because of same maddening at our side. You have very solid questions that we are enthusiastic to tell more and more: 1) Automatic instrumentation is what creates no code change. Once you add Thundra to your environment variables, you can change the monitoring settings with annotation or additional environment variables with no code change for Java. For Node.js and Go, You simply wrap your functions with our agents. With manual instrumentation on the other hand, you can also add code blocks to inspect your variables. You can check our codes here: https://github.com/thundra-io https://github.com/thundra-io. We are open to share more, if you want any further questions. 2) Zero overhead is one of other strong points of us. You need to switch to async. monitoring for this. This means you need to add "our" lambda to your environment variables (Please check for it: https://github.com/thundra-io/serverless-plugin-thundra-monitor-cw https://github.com/thundra-io/serverless-plugin-thundra-moni...). Then this lambda sends logs of your function to us. Overhead of sending and retrying for failed logs will not happen. Only overhead can occur because of make your code to gather more logs for our lambda to read. But this is truly negligible.
bap 9y ago
This is very timely for me and Good Work! Now PLEASE tell me your logo will be a cat?? (Yes I can see that you seem to going for 'bird' sigh)
- emrahsamdan 9y ago
  Sorry for make you sad but Thundra is a bird. Inspired by Thundra (a.k.a queen of rainforest :)) in Aladdin: http://disney.wikia.com/wiki/Thundra http://disney.wikia.com/wiki/Thundra.
emrahsamdan 9y ago
Hey! We have made our first webinar! It includes a demo showing solving a real life problem in Java with RDS. You can watch it here: https://www.youtube.com/watch?v=bmAw24uquK0 https://www.youtube.com/watch?v=bmAw24uquK0

74 points by CSDude 9y ago

This looks promising! We tried using a similar service "IOPipe" last year and really had trouble with their offering. https://www.iopipe.com/ https://www.iopipe.com/ It would be nice to see more companies providing this kind of observability services ( especially for multiple cloud providers )

sozal 9y ago
Hi, Thundra has more advanced features such as - Automated instrumentation for tracing (no code change is required) - Aggregating metrics and logs with traces - Async monitoring over CloudWatch logs with almost zero overhead - Distributed tracing support by AWS X-Ray integration - and others ... Check www.thundra.io out for more
dankohn1 9y ago
CNCF is current tracking 8 projects/products in the serverless tools space, including Thundra and IOPipe: https://landscape.cncf.io/grouping=landscape&landscape=tools https://landscape.cncf.io/grouping=landscape&landscape=tools

bthornbury 9y ago

As an interested user, what is the difference here between traditional instrumentation like Application Insights? I understand this is intended for AWS Lambda, but what is it doing that makes it more suited?

sozal 9y ago
Hi, Instrumentation (manual and automated) for tracing is just one of the features of Thundra. Thundra provides all the three pillars of observability in one tool. It can aggreagate metrics and logs. Thundra is also integrated with AWS X-Ray for supporting distributed tracing. Additionally, Thundra can monitor your Lambda functions with almost zero overhead by its async monitoring functionality over CloudWatch logs.

jedberg 9y ago

I tried creating an account but it just says: "Error: server_error. Check the console for further details." :( This looks really interesting! Also, the title here says "Python", but I don't see Python in the docs. I assume it is similar to Go, where you just import a library?

sozal 9y ago
Hi, Can you try by enabling 3rd party cookies or whitelist Auth0's domain in your browser?
- jedberg 9y ago
  I switched browsers from Safari to Chrome and it worked. Any info on the Python question? Thanks!
  - sozal 9y ago
    Unfortunately Python is not supported yet but it is in progress. It will be available in a couple of weeks. Thanks for your interest to Thundra

reconbot 9y ago

I've built a cloudwatch metrics stack that reads out of lambda's logs. There's no other way to get data out of your lambda for "free". XRAY is an option but not everything you'd want.

jelder 9y ago
I went through the trouble of setting up X-Ray for three of our Serverless projects (~50 Lambda functions) and can't say that we've gotten much value from it. The UI is... lacking.
- emrahsamdan 9y ago
  Thanks for giving a try! We are here to continuously improve our product with your feedbacks. What specifically can you recommend to improve our UI?

thinkingkong 9y ago

This is cool! Strategically are you worried about being built on top of an AWS specific solution? The only reason I ask is its pretty inevitable that AWS will be providing instrumentation / apm capabilities native to their offering.

sozal 9y ago
Hi, Thundra is not designed just for AWS Lambda. Even though currently we are supporting it on AWS Lambda platform, it will be available for other environments such as containers soon.

adevine 9y ago

I'm not quite sure I understand what the log support entails. Also, it says for Node.js that this is coming soon. Any idea of the ETA on this one?

berkay 9y ago
You can select a method and see the logs generated for that method. Idea is to have all the information - request parameters, return values, logs, metrics all in one place to make it easier to troubleshoot problems.

biztos 9y ago

This would be super timely for me if it supported Python, but even without that it's interesting. Coming from the world of servers (virtual or physical) I find the logging latency and the lack of "observability" in general quite maddening in Lambda. Two things I wonder about this: 1) The "no code changes" part implies it's adding another layer of code somewhere, which means another thing to trust and probably not be able to audit, which is not automatically OK outside the "move fast break things" world. Am I misreading it? 2) How can you add metrics with "no overhead?" Surely there must be some overhead... and then it would be nice to be able to measure how much. For me personally, #1 is a real issue and #2 is just hypothetical, but I would expect it to be the other way around for a fast-growing startup.

adjohn 9y ago
If you're in a rush, IOpipe[1] already has Python support offering profiling, tracing, and other debugging and observability tools for AWS Lambda. Co-founder of IOpipe here, feel free to ask anything. [1] - https://www.iopipe.com https://www.iopipe.com
emrahsamdan 9y ago
You are right about lack of observability tools for AWS Lambda. Thundra was born because of same maddening at our side. You have very solid questions that we are enthusiastic to tell more and more: 1) Automatic instrumentation is what creates no code change. Once you add Thundra to your environment variables, you can change the monitoring settings with annotation or additional environment variables with no code change for Java. For Node.js and Go, You simply wrap your functions with our agents. With manual instrumentation on the other hand, you can also add code blocks to inspect your variables. You can check our codes here: https://github.com/thundra-io https://github.com/thundra-io. We are open to share more, if you want any further questions. 2) Zero overhead is one of other strong points of us. You need to switch to async. monitoring for this. This means you need to add "our" lambda to your environment variables (Please check for it: https://github.com/thundra-io/serverless-plugin-thundra-monitor-cw https://github.com/thundra-io/serverless-plugin-thundra-moni...). Then this lambda sends logs of your function to us. Overhead of sending and retrying for failed logs will not happen. Only overhead can occur because of make your code to gather more logs for our lambda to read. But this is truly negligible.

bap 9y ago

This is very timely for me and Good Work! Now PLEASE tell me your logo will be a cat?? (Yes I can see that you seem to going for 'bird' sigh)

emrahsamdan 9y ago
Sorry for make you sad but Thundra is a bird. Inspired by Thundra (a.k.a queen of rainforest :)) in Aladdin: http://disney.wikia.com/wiki/Thundra http://disney.wikia.com/wiki/Thundra.

emrahsamdan 9y ago

Hey! We have made our first webinar! It includes a demo showing solving a real life problem in Java with RDS. You can watch it here: https://www.youtube.com/watch?v=bmAw24uquK0 https://www.youtube.com/watch?v=bmAw24uquK0