Welcome!

Blog Feed Post

Understanding the G1 Garbage Collector – Java 9

Oracle’s Java 9 Hotspot VM ships with the Garbage First (G1) GC as its default garbage collector. This GC, first introduced in Java 7, has the unique ability to efficiently and concurrently deal with very large heaps. It can also be configured to not exceed a maximum pause time. In this post we’ll take a look at how the G1 works compared to other collectors and why it can so easily outperform other state-of-the-art GCs on large heaps.

Most state-of-the-art GCs classify heaps into either young generation or old generation objects. This is done mostly because studies of real-world Java applications have shown that more than 90% of objects don’t survive their first garbage collection. Older objects (objects that have survived a few collection cycles) however tend to remain alive and have a 98% chance of surviving. Java GCs split the young generation objects further into the survivor space and the Eden space. Newly allocated objects are always allocated to the Eden space. Once an object survives its first garbage collection, it’s moved to the survivor space. When an object survives multiple collection cycles, it’s eventually moved to the older generation. This is done so that a run-time efficient algorithm can be used on new objects (this algorithm’s run-time depends only on the number of surviving objects but wastes half the heap size) and a memory-efficient algorithm can be used on the old generation (this algorithm’s run-time depends on the heap size, but it uses available memory as efficiently as possible). A heap of such a collector could look like this:

Compared to most other garbage collectors, the G1 has two big advantages: (1) it can do most of its work concurrently (i.e., without halting application threads), and (2) it uses non-continuous spaces, which enables the G1 to efficiently deal with very large heaps. Also, the G1 can collect both the young and the old generation at once. This has something to do with the unique way the G1 uses the available heap. Instead of splitting the heap into three spaces (Eden, survivor, and old) like most other GCs, it splits the heap into many (often several hundred) very small regions. These regions are fix-sized (about 2Mb by default). Each region is assigned to a space.  A G1 heap could look like this:


(O: old, S: survivor, E: Eden, U: unassigned)

Splitting the heap into small regions enables the G1 to select a small group of regions to collect and finish quickly. If a region is scheduled for collection, all surviving objects will be copied from the collected region to an unassigned region. Assuming that the collected region was of the Eden space, the unassigned region holding all surviving objects will then become a survivor region. Ideally, if a region is full of garbage (meaning it doesn’t contain a single surviving object), the region can be declared “unassigned” and no work has to be done there.

Granted, if one wants to collect the entire heap, the G1 has to do the same amount of work as any other GC, but this is where the G1 shines because it doesn’t have to collect the entire heap. It doesn’t even have to collect an entire generation. It can select any number or combination of regions to collect. To optimize collection time, it always selects regions that are full (or almost full) of garbage and thereby minimizes the amount of work it has to do to free heap space for subsequent allocations. Other GCs always collect an entire generation, meaning their run-time complexity often depends on the total heap size. In the G1 case however, this depends on the amount of live objects because memory can be freed without handling an entire generation. Ideally, when the heap is big enough, some regions will always be completely full of garbage, making it easy to collect them.

In addition, the G1 can do most of its work concurrently. In the Java world, we already know about concurrent collections from the Concurrent Mark & Sweep GC (CMS). However, the CMS can only collect the old generation concurrently, it still needs to halt the application to collect the young generation. The G1 only stops the application at the beginning of the GC to do some quick bookkeeping before it immediately resumes the application. This phase is called the “Initial Mark”. Then, while the application is executing, the GC will follow all references and mark life objects (“Concurrent Mark” phase). When this is done, the application is suspended again, and a final cleanup is made (“Final Mark” phase) before selecting a few regions and collecting them (“Evacuation” phase). As the evacuation phase is fast, especially for large heaps, the G1 usually outperforms other GCs in terms of suspension time of the executed application.

The downside is that the G1 doesn’t perform well with small heaps. Where the sweet spot lies depends on the application in question. If needed, you can always fall back to the old default collector by setting the -XX:+UseParallelOldGC flag. If the G1 has too little heap available, you will see “Full GC”s in the GC log. A Full GC is only performed when the G1 determines that the usual mode of operation as described above is no longer possible. In this case, the entire heap is collected with a memory-efficient (but slow) algorithm that ideally will make things better for the next collection. If there are “Full GC”s occurring, you should increase the heap size if possible and fall back to another GC if necessary.

Due to the small regions, the G1 can be configured to limit its maximum pause time by setting -XX:MaxGCPauseMillis=n. The G1 will then estimate the maximum number of regions it can collect at once without overstepping this limit based on previous collections as well as on the amount of detected garbage. The G1 is still far from being a real-time collector, however, it performs better than other collectors that can’t even begin to adhere to such a limit because of their rigid heap structures.

The post Understanding the G1 Garbage Collector – Java 9 appeared first on Dynatrace blog – monitoring redefined.

Read the original blog entry...

More Stories By APM Blog

APM: It’s all about application performance, scalability, and architecture: best practices, lifecycle and DevOps, mobile and web, enterprise, user experience

Latest Stories
DX World EXPO, LLC, a Lighthouse Point, Florida-based startup trade show producer and the creator of "DXWorldEXPO® - Digital Transformation Conference & Expo" has announced its executive management team. The team is headed by Levent Selamoglu, who has been named CEO. "Now is the time for a truly global DX event, to bring together the leading minds from the technology world in a conversation about Digital Transformation," he said in making the announcement.
"Space Monkey by Vivent Smart Home is a product that is a distributed cloud-based edge storage network. Vivent Smart Home, our parent company, is a smart home provider that places a lot of hard drives across homes in North America," explained JT Olds, Director of Engineering, and Brandon Crowfeather, Product Manager, at Vivint Smart Home, in this SYS-CON.tv interview at @ThingsExpo, held Oct 31 – Nov 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA.
SYS-CON Events announced today that Conference Guru has been named “Media Sponsor” of the 22nd International Cloud Expo, which will take place on June 5-7, 2018, at the Javits Center in New York, NY. A valuable conference experience generates new contacts, sales leads, potential strategic partners and potential investors; helps gather competitive intelligence and even provides inspiration for new products and services. Conference Guru works with conference organizers to pass great deals to gre...
DevOps is under attack because developers don’t want to mess with infrastructure. They will happily own their code into production, but want to use platforms instead of raw automation. That’s changing the landscape that we understand as DevOps with both architecture concepts (CloudNative) and process redefinition (SRE). Rob Hirschfeld’s recent work in Kubernetes operations has led to the conclusion that containers and related platforms have changed the way we should be thinking about DevOps and...
The Internet of Things will challenge the status quo of how IT and development organizations operate. Or will it? Certainly the fog layer of IoT requires special insights about data ontology, security and transactional integrity. But the developmental challenges are the same: People, Process and Platform. In his session at @ThingsExpo, Craig Sproule, CEO of Metavine, demonstrated how to move beyond today's coding paradigm and shared the must-have mindsets for removing complexity from the develop...
In his Opening Keynote at 21st Cloud Expo, John Considine, General Manager of IBM Cloud Infrastructure, led attendees through the exciting evolution of the cloud. He looked at this major disruption from the perspective of technology, business models, and what this means for enterprises of all sizes. John Considine is General Manager of Cloud Infrastructure Services at IBM. In that role he is responsible for leading IBM’s public cloud infrastructure including strategy, development, and offering m...
The next XaaS is CICDaaS. Why? Because CICD saves developers a huge amount of time. CD is an especially great option for projects that require multiple and frequent contributions to be integrated. But… securing CICD best practices is an emerging, essential, yet little understood practice for DevOps teams and their Cloud Service Providers. The only way to get CICD to work in a highly secure environment takes collaboration, patience and persistence. Building CICD in the cloud requires rigorous ar...
Companies are harnessing data in ways we once associated with science fiction. Analysts have access to a plethora of visualization and reporting tools, but considering the vast amount of data businesses collect and limitations of CPUs, end users are forced to design their structures and systems with limitations. Until now. As the cloud toolkit to analyze data has evolved, GPUs have stepped in to massively parallel SQL, visualization and machine learning.
"Evatronix provides design services to companies that need to integrate the IoT technology in their products but they don't necessarily have the expertise, knowledge and design team to do so," explained Adam Morawiec, VP of Business Development at Evatronix, in this SYS-CON.tv interview at @ThingsExpo, held Oct 31 – Nov 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA.
To get the most out of their data, successful companies are not focusing on queries and data lakes, they are actively integrating analytics into their operations with a data-first application development approach. Real-time adjustments to improve revenues, reduce costs, or mitigate risk rely on applications that minimize latency on a variety of data sources. In his session at @BigDataExpo, Jack Norris, Senior Vice President, Data and Applications at MapR Technologies, reviewed best practices to ...
Widespread fragmentation is stalling the growth of the IIoT and making it difficult for partners to work together. The number of software platforms, apps, hardware and connectivity standards is creating paralysis among businesses that are afraid of being locked into a solution. EdgeX Foundry is unifying the community around a common IoT edge framework and an ecosystem of interoperable components.
"ZeroStack is a startup in Silicon Valley. We're solving a very interesting problem around bringing public cloud convenience with private cloud control for enterprises and mid-size companies," explained Kamesh Pemmaraju, VP of Product Management at ZeroStack, in this SYS-CON.tv interview at 21st Cloud Expo, held Oct 31 – Nov 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA.
Large industrial manufacturing organizations are adopting the agile principles of cloud software companies. The industrial manufacturing development process has not scaled over time. Now that design CAD teams are geographically distributed, centralizing their work is key. With large multi-gigabyte projects, outdated tools have stifled industrial team agility, time-to-market milestones, and impacted P&L stakeholders.
"Akvelon is a software development company and we also provide consultancy services to folks who are looking to scale or accelerate their engineering roadmaps," explained Jeremiah Mothersell, Marketing Manager at Akvelon, in this SYS-CON.tv interview at 21st Cloud Expo, held Oct 31 – Nov 2, 2017, at the Santa Clara Convention Center in Santa Clara, CA.
Enterprises are adopting Kubernetes to accelerate the development and the delivery of cloud-native applications. However, sharing a Kubernetes cluster between members of the same team can be challenging. And, sharing clusters across multiple teams is even harder. Kubernetes offers several constructs to help implement segmentation and isolation. However, these primitives can be complex to understand and apply. As a result, it’s becoming common for enterprises to end up with several clusters. Thi...