Using Dynatrace AI based Monitoring to Ensure Blockchain Vertoning, Dynatrace blog
The explosion of Bitcoin and cryptocurrency has bot a proving ground for Blockchain transactions. Blockchain is a permanently growing list of records called blocks, which are connected and secured using cryptography (1). Blockchains provide voortdurend verifiable means of recording transactions inbetween two parties. Serious actors te Finance, eCommerce and IT are massively investing ter and building out Blockchain frameworks. Blockchain by its very nature requires tremendous processing power and network conversations with actors exchanging blocks. There has not bot a lotsbestemming of research into the spectacle influence adding Blockchain transactions into ecommerce applications will have. Blockchain adoption raises several operational concerns:
- It is unclear how poor vertoning will influence end users or applications relying on a Blockchain transaction.
- Spil Blockchain becomes part of the fabric of cloud based, fresh stack ecommerce/financial applications the onus on how Blockchain behaves and performs will not only fall on to those providing the core Blockchain processing, but any actor integrating it into their applications or exchanges.
- The explosion of complexity around cloud and container based approaches thesis Blockchain applications will be based on will require fresh discipline and insight to ensure that they are adequately monitored from a spectacle perspective.
That last voorwerp is where most organizations will fight. Blockchain use at web scale will toebijten te ridiculously ingewikkeld environments which will include elastically scaling compute resources which may only exist temporarily. The scale of thesis applications will require a fresh monitoring paradigm spil it will be impractical and te many cases unlikely for humans to monitor the health of thesis Blockchain components.
Most literature which can be found on monitoring Blockchain concentrates on very specific metrics regarding the health of the chain. Here is a view from an IBM Bluemix UI.
Above you can see metrics like Total Blocks, Latest Instantiations, Invocations, etc. (Two)
Below is an example of some Blockchain activity from https://Blockchain.informatie/charts
Above you can see that te this example Blockchain gegevens is being aggregated into Block Details, Mining Information, Network Activity and Wallet Activity. This is clearly pertinent information about the Blockchain, however:
- What happens if there is an infrastructure punt with your Blockchain processing?
- What happens if there is a network kwestie inbetween Blockchain actors?
- How do the above metrics permit an organization to diagnose that a spectacle kwestie is occurring with their Blockchain?
- How do the above metrics permit an organization to understand the influence to end users or service calls depending on a Blockchain?
Monitoring Blockchains requires visibility into the entire technology stack. Te addition, monitoring Blockchains requires monitoring every transaction. You can’t skip requests or sample/throttle gegevens when monitoring Blockchains and blindly trust that the application, services, process, network or infrastructure layers are always providing 100% availability and optimal spectacle.
Automating Spectacle Monitoring for Blockchains
Here is an example of how to monitor a ordinary Blockchain based on Ethereum (Trio). Ethereum (https://www.ethereum.org) is an open-source Blockchain technology. Ter this case it is deployed spil GOlang (a Google programming language popular ter fresh stack deployments).
Te this example, wij are using Dynatrace to automatically detect the GO processes which were created when wij spun up Ethereum ter the Amazon Cloud using EC2 (AWS).
The Blockchain is being computed te the geth process.
Spil Blockchains get deployed into containers and micro-services te cloud environments, Dynatrace automatically detects thesis processes. Dynatrace will automatically pull te the loom files from the geth process (more on this te a bit).
Blockchains are about communicating inbetween two parties, below is how Dynatrace automatically detects requests being made to potential externally held Blockchains. Calls to Blockchains can be executed ter a diversity of ways, one of which is via JSON based Surplus calls.
The screenshot above shows how outward request to Blockchains can be monitored. Below is an example of how network traffic coming ter to and out of our cloud example hosting the Blockchain is discovered and monitored.
By including Dynatrace spil part of your container/cloud deployments you can automate the discovery of your Blockchain services. This is spil plain spil including Dynatrace te your buildpacks, spil your Blockchain compute resources are created they will be automatically instrumented and monitored.
Deterministic AI to the Rescue
The sample Blockchain I demonstrated above is an overly ordinary example. Blockchains will exist ter very elaborate environments executing thousands of transactions vanaf minute.
Thesis environments will challenge even the most mature IT organization to monitor thesis business-critical transactions. This is where deterministic Artificial Intelligence monitoring, using specific machine learning algorithms will be able to auto-discover, auto-baseline 100% of your traffic ter a total stack style (from End User to Application to Blockchain to Infrastructure) and be able to determine the influence of spectacle events spil well spil the root cause.
Deterministic AI is the only treatment to effectively monitor thesis environments. Cognitive AI will take too long to understand all of the entities involved, and by the time it could have understood the relationships the environment will have switched. Monitoring Blockchain containers running ter elastic cloud environments which will be permanently switching and evolving requires a deterministic AI treatment. This is what Dynatrace provides, an inherent understanding of all the compute entities involved ter operating a Blockchain based application. Understanding the entities, relationships and dependencies is only part of what needs to be done. Dynatrace employs a diversity of machine learning algorithms to baseline, compare time frames, create multi-dimensional views of gegevens to then determine and rank issues which inevitably occur te the complicated environments.
What this provides organizations is the capability to build self-healing Blockchain applications. For example, if a Blockchain process is having a memory punt, Dynatrace can detect that state and be used to trigger a Chef, Puppet or Ansible script to adjust the memory settings automatically to maintain that Blockchain processing.
Tailor Your Blockchain Monitoring
While the Dynatrace AI will automatically detect and baseline your Blockchain processes, there may be situations where there are details locked away ter your Blockchain logs that you want to be aware of.
Below is an example of how Dynatrace is able to monitor Blockchain loom files. You can create rules to look for significant strings within your Blockchain loom files and automatically trigger outward events based on those events/strings.
This works automatically and doesn’t require expensive, complicated third party loom analysis implements pushing gegevens into third party gegevens aggregation contraptions which would then have to thrust out to outer automation scripts.
Te addition to monitoring Blockchain loom files, Dynatrace can be used to insert extra metadata, what wij call tags, overheen your Blockchain compute resources. Thesis tags could be used to pull things like the Chaincode ID, Chaincode Version or Peer ID. This will permit your teams to create separate visualizations, reports, alerts, etc… for only those Blockchains they are working with.
Blockchain is Too Significant Not to Monitor, Attempt it for Yourself!
The example wij displayed above is a highly-simplified example, your Blockchain applications will exist ter much more complicated environments. Wij invite organizations to attempt Dynatrace te your Blockchain enabled environments. Here is a Free Trial of Dynatrace which you can download and install into your environments where you are running your Blockchain processes. Dynatrace can lightly install on to Hosts directly or you can include Dynatrace ter your container/cloud build-packs with nominal effort.
Go ahead and give this a attempt with your Blockchains, please postbode your practice below, wij would love to get your terugkoppeling.
David Jones is the Director of Sales Engineering and AIOPs Evangelism for Dynatrace. He has bot with Dynatrace for Ten years, and has 20 years’ practice working with web and mobile technologies from the very first commercial HTML editor to the latest web delivery platforms and architectures. He has worked with scores of Fortune 500 organizations providing them the most latest industry best practices for web and mobile application delivery. Prior to Dynatrace he has worked at Gomez (Waltham), S1 Corp (Atlanta), Broadvision (Bay Area), Interleaf/Texcel (Waltham), i4i (Toronto) and SoftQuad (Toronto).