[ad_1]
CEO of lookoutOver the last 20 years, he has led data-driven business, product and R&D teams from startups to government agencies.
Getty
Businesses across all industries are preparing for a recession by cutting non-essential spending. However, our investments in technology and modernization have remained consistent and continue to grow. From our perspective, this is a smart decision that highlights the fact that investments in IT and better tools lead to lower costs and higher ROI.
That’s not to say that big company deals aren’t taken seriously when it’s time to renew. If you’re not building a positive relationship with your customers as a technology vendor, don’t be surprised if your churn rate is higher than normal as companies strive to invest prudently. But the problem with smart technology spending that more and more CFOs are beginning to understand (hence the rise of FinOps) is that the new capabilities and productivity gains unlocked by better technology greatly outweigh cost concerns. am. Valuable software vendors can easily demonstrate their value.
And I can’t think of any easier way for companies to avoid unnecessary costs and improve productivity than adopting dynamic logging.
The traditional world of logging
In an ideal world, developers should not have to deal with the cost of storing logs, both from a financial and application performance standpoint, and should not waste any valuable time creating log lines. is not. Instead, developers can easily extract the data they need on demand. But for most companies, logging hasn’t changed in decades.
So what do many of us do? As developers, we need to debug a problem and are afraid of situations where we don’t have the logs we need, so we insert log lines every few lines of code doing. This is a universal concern, so unfortunately we tend to log everything. However, the fact remains that logs quickly become expensive to store and noisy at scale.
Solution: dynamic logging
The idea behind dynamic logging is that you can change the verbosity of your logs on the fly without stopping your application or writing new code. This means that the logs can be left in a low verbosity state when there are no issues, and when an issue arises that requires troubleshooting, the developer can increase the verbosity and click through to all the data they need. can only be collected. button.
There are several dynamic logging solutions on the market. Our company has his Live Logger product, IBM has its own dynamic logging solution, and Pixie Labs is currently working on a product (alpha) that addresses the problem of converting static logs to dynamic logs in Go. is. And the traditional logging player itself will reveal a more dynamic solution.
Unfortunately, we are not yet at the stage where logs can be completely eliminated. But dynamic logging is a step in the right direction to reduce noise, reduce costs, and improve productivity.
So how does it work?
Each log entry is categorized by the following severity classes and debug components, listed in order of least impact to most impact.
logging category
- trace
- debug
- information
- Warning
- error
- Deadly
You can dynamically change the log class without stopping your application. This affects the verbosity that log lines produce. Traditionally this could be considered “noise” when the log lines are very detailed but irrelevant to the problem at hand, but with dynamic logging you can dynamically change classes You can leave the log in a permanent Trace or Debug state until you choose not to. It will be an error or fatal on the fly.
This dynamic instrumentation is made possible through bytecode manipulation. This is commonly seen in cybersecurity, but not common in developer tools. Combined with modern debugging, these tools give developers instant access to the code-level data they need to troubleshoot and understand complex modern applications.
What to Look for in a Dynamic Logging Solution
1. Ability to change log verbosity without it Need to write new code or restart the application. If this is not possible, the solution is not dynamic and does not provide real-time functionality.
2). Ability to dynamically search and filter missing log data. This allows engineering teams to zoom in on application components and change the redundancy of specific files, functions, and services.
3. Tight integration with existing logging tools. In the future all logging tools will have dynamic capabilities, but this is not the current state of the world. Abandoning and replacing legacy logging solutions already in place is a big undertaking. Next-generation logging solutions can aggregate existing logging data and make it more dynamic and actionable by providing seamless integration.
Migrating to dynamic logging is straightforward for most organizations. Optimize costs, reduce noise, and make life easier for engineers. Get the data you need, when you need it. No more overwhelmed with logs and searching for needles in haystacks when troubleshooting.
The Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. am i eligible?
[ad_2]
Source link
