Bigtable is a scalable, distributed, high-performance NoSQL database that processes more than 6 billion requests per second at peak and has more than 10 Exabytes of data under management. Operating at this scale, Bigtable is highly optimized for high-throughput and low-latency reads and writes. Even so, our performance engineering team continually explores new areas to optimize. In this article, we share details of recent projects that helped us push Bigtable’s performance envelope forward, improving single-row read throughput by 20-50% while maintaining the same low latency.

Below is an example of the impact we delivered to one of our customers, Snap. The compute cost for this small-point read-heavy workload reduced by 25% while maintaining the previous level of performance.

Performance research

We use a suite of benchmarks to continuously evaluate Bigtable’s performance. These represent a broad spectrum of workloads, access patterns and data volumes that we see across the fleet. Benchmark results give us a high-level view of performance opportunities, which we then enhance using sampling profilers and pprof for analysis. This analysis plus several iterations of prototyping confirmed feasibility of improvements in the following areas: Bloom filters, prefetching, and a new post-link-time optimization framework, Propeller.

Bloom filters

Bigtable stores its data in a log-structured merge tree. Data is organized into row ranges and each row range is represented by a set of SSTables. Each SSTable is a file that contains sorted key-value pairs. During a point-read operation, Bigtable searches across the set of SSTables to find data blocks that contain values relevant to the row-key. This is where the Bloom filter comes into play. A Bloom filter is a space-efficient probabilistic data structure that can tell whether an item is in a set, it has a small number of false positives (item may be in the set), but no false negatives (item is definitely not in the set). In Bigtable’s case, Bloom filters reduce the search area to a subset of SSTables that may contain data for a given row-key, reducing costly disk access.

We identified two major opportunities with the existing implementation: improving utilization and reducing CPU overhead.

First, our statistics indicated that we were using Bloom filters in a lower than expected percentage of requests. This was due to our Bloom filter implementation expecting both the “column family” and the “column” in the read filter, while a high percentage of customers filter by “column family” only — which means the Bloom filter can’t be used. We increased utilization by implementing a hybrid Bloom filter that was applicable in both cases, resulting in a 4x increase in utilization. While this change made the Bloom filters larger, the overall disk footprint increased by only a fraction of a percent, as Bloom filters are typically two orders of magnitude smaller than the data they represent.

Second, the CPU cost of accessing the Bloom filters was high, so we made enhancements to Bloom filters that optimize runtime performance: 

• Local cache for individual reads: When queries select multiple column families and columns in a single row, it is common that the query will use the same Bloom filter. We take advantage of this by storing a local cache of the Bloom filters used for the query being executed.
• Bloom filter index cache: Since Bloom filters are stored as data, accessing them for the first time involves fetching three blocks — two index blocks and a data block — then performing a binary search on all three. To avoid this overhead we built a custom in-memory index for just the Bloom filters. This cache tracks which Bloom filters we have in our block cache and provides direct access to them.

Overall these changes decreased the CPU cost of accessing Bloom filters by 60-70%.

Prefetching

In the previous section we noted that data for a single row may be stored in multiple SSTables. Row data from these SSTables is merged into a final result set, and because blocks can either be in memory or on disk, there’s a risk of introducing additional latency from filesystem access. Bigtable’s prefetcher was designed to read ahead of the merge logic and pull in data from disk for all SSTables in parallel.

Prefetching has an associated CPU cost due to the additional threading and synchronization overhead. We reduced these costs by optimizing the prefetch threads through improved coordination with the block cache. Overall this reduced the prefetching CPU costs by almost 50%.

Post-link-time optimization

Bigtable uses profile guided optimizations (PGO) and link-time optimizations (ThinLTO). Propeller is a new post-link optimization framework released by Google that improves CPU utilization by 2-6% on top of existing optimizations.

Propeller requires additional build stages to optimize the binary. We start by building a fully optimized and annotated binary that holds additional profile mapping metadata. Then, using this annotated binary, we collected hardware profiles by running a set of training workloads that exercise critical code paths. Finally, using these profiles as input, Propeller builds a new binary with an optimized and improved code layout. Here is an example of the improved code locality.

The new build process used our existing performance benchmark suite as a training workload for profile collection. The Propeller optimized binary showed promising results in our tests, showing up to 10% improvement in QPS over baseline. 

However, when we released this binary to our pilot production clusters, the results were mixed. It turned out that there was overfitting for the benchmarks. We investigated sources of regression by quantifying profile overlap, inspecting hardware performance counter metrics and applied statistical analysis for noisy scenarios. To reduce overfitting, we extended our training workloads to cover a larger and more representative set of use cases. 

The result was a significant improvement in CPU efficiency — reducing fleetwide utilization by 3% with an even more pronounced reduction in read-heavy workloads, where we saw up to a 10% reduction in CPU usage.

Conclusion

Overall, single-row read throughput increased by 20-50% whilst maintaining the same latency profile. We are excited about these performance gains, and continue to work on improving the performance of Bigtable. Click here to learn more about Bigtable performance and tips for testing and troubleshooting any performance issues you may encounter.

The last few years have underscored the importance of speed in bringing new drugs and medical devices to market, while ensuring safety and efficacy. Over this time, healthcare and life sciences organizations have transformed the way they research, develop, and deliver patient care by embracing agility and innovation.

Now, the industry is set to reap the benefits of cloud technology and overcome the existing barriers to innovation.

What’s holding back innovation?

Costly clinical trials: The process of trialing and developing new drugs and devices is still long and costly, with more than 1 in 5 clinical trials failing due to a lack of funding.1 The high failure rate comes as no surprise when you consider the average clinical trial costs $19 million and takes 10-15 years (through all 3 phases) to be approved.2

Stringent security requirements: Pre-clinical R&D and clinical trials use large volumes of highly sensitive patient data – making the life sciences industry one of the top sectors targeted by hackers.3 On top of this, the FDA and other regulatory bodies have strict requirements for medical device cybersecurity.

Unpredictable supply chains: Global supply chains are becoming increasingly complex and unpredictable. This can be brought on by anything from supply shortages, to geo-political events, and even bad weather. Making things worse is the lack of visibility into medical shipment disruptions – so when disaster strikes you’re often caught off guard.

Google Cloud for life sciences

At Alphabet, we’ve made significant investments in healthcare and life sciences, helping to tackle the world’s biggest healthcare problems, from chronic disease management, to precision medicine, to protein folding.

Together with Google, you can transform your life sciences organization and deliver secure, data-driven innovation across the value chain.

• Accelerate clinical trials to deliver life-saving treatments faster and at less cost. Clinical trials require relevant and equitable patient cohorts that can produce clinically valid data. Solutions like DocAI can enable optimal patient matching for clinical trials, helping organizations optimize clinical trial selection and increase time to value. How that patient data is collected is also important. Collection in a physician’s office captures a snapshot of the participant’s data at one point in time and doesn’t necessarily account for daily lifestyle variables. Fitbit, used in more than 1,500 published studies–more than any other wearable device–can enrich clinical trial endpoints with new insights from longitudinal lifestyle data, which can help improve patient retention and compliance with study protocols. We have introduced Device Connect for Fitbit, which empowers healthcare and life sciences enterprises with accelerated analytics and insights to help people live healthier lives. We are able to empower organizations to improve clinical trials in key ways:

1. Enable clinical trial managers to quickly create and launch mobile and web RWE collection mechanism for patient reported outcomes
2. Enable privacy controls with Cloud Healthcare Consent API and, as needed, remove PHI using Cloud Healthcare De-identification API
3. Ingest RWE and data into BigQuery for analysis
4. Leverage Looker to enable quick visualization and powerful analysis of a study’s progress and results

• Ensure security and privacy for a safe, coordinated, and compliant approach to digital transformation. Google Cloud offers customers a comprehensive set of services including pioneering capabilities such as BeyondCorp Enterprise for Zero Trust and VirusTotal for malicious content and software vulnerabilities; Chronicle’s security analytics and automation coupled with services such as Security Command Center to help organizations detect and protect themselves from cyber threats; as well as expertise from Google Cloud’s Cybersecurity Action Team. Google Cloud also recently acquired Mandiant, a leader in dynamic cyber defense, threat intelligence and incident response services.

• Optimize supply chains and enhance your data to prepare for the unpredictable. With a digital supply chain platform, we can empower supply chain professionals to solve problems in real time including visibility and advanced analytics, alert-based event management, collaboration between teams and partners, and AI-driven optimization and simulation.

Ready to learn more? We’ll be taking a deep dive into each of the challenges outlined above in our life sciences video series. Stay tuned.

1. National Library of Medicine
2. How much does a clinical trial cost?
3. Life Sciences Industry Becomes Latest Arena in Hackers’ Digital Warfare

May is Asian American Pacific Islander Heritage Month — a time for us to come together to celebrate and remember the important people and history of Asian and Pacific Island heritage. This feature highlights founder Georgene Huang and how her startup Fairygodboss uses easy to use Google Cloud tools to grow and innovate. 

Coming from a male-dominated industry like Wall Street, I never thought much about my gender. I was so focused on work that I didn’t feel like I could focus on my identity as a woman of color. That all changed when I found myself thrust into the job market while two months pregnant.

Traditionally, job-seeking sites are built by – and geared towards – men and their professional needs. There was no recruiting site that broke down what candidates who identify as women might expect from a company, let alone women of color like myself. I founded Fairygodboss to create a space where women can crowdsource information about opportunities and employers in order to make informed career decisions. 

Fairygodboss 

Fair•y•god•boss (noun): A person who elevates women at work.

Fairygodboss is the largest career community for women. In researching potential employers, I had many questions surrounding company culture, maternity leave, and women in leadership positions. While sites like Glassdoor provide high-level details for job seekers, I wanted more concrete examples and data than general salary information and anecdotes about work-life balance. 

Fairygodboss began as a job review site geared towards women, but has expanded significantly since our initial launch. In addition to providing free resources like career connections, job listings, virtual recruiting events, community advice, and real research on how companies treat women, we also proactively help companies improve gender diversity. Fairygodboss partners with hundreds of top employers to create a 360-degree recruiting and employer branding program designed to drive female applicants to job postings by showcasing a company’s track record of women in leadership within a company, and their diversity investments. 

Google Cloud + Fairygodboss: Data defined 

At our core, Fairygodboss is a data-driven company, and we rely on Google Cloud to help us process, store, and analyze user-driven events occurring on the site. As a highly scalable, serverless, and cost-effective data warehousing solution, BigQuery collect millions of events each day, helping us identify where traffic is coming from (and where it goes next), and it produces actionable insights to provide better recommendations, information, and opportunities for women visiting our platform. 

After seamlessly connecting our BigQuery data to Data Studio, we have been able to visualize and make sense of all the data we receive on our platform. Unlike some other data visualization tools, Data Studio allows us to analyze data in real time. With tools like BigQuery, we are now able to identify trends and gain a deep understanding for how our users interact with our site. 

Google for Startups Accelerator: Women Founders 

In order to provide more accurate job recommendations for our users, the Fairygodboss team wanted to build a machine learning prototype to classify our users and jobs into categories to facilitate more effective matching. So we applied for the Google for Startups Accelerator: Women Founders, a three-month digital accelerator program for high-potential Seed to Series A tech startups based in the U.S. and Canada. Along with tailored mentorship and product support from Googlers and industry experts, the Google for Startups Accelerator provided $100K in Google Cloud credits to scale our businesses. The mentorship we received from the Google Cloud technical experts as part of the Accelerator – special shout out to Peter Novig! – empowered us to integrate Cloud AutoML into our systems and ultimately curate more accurate job suggestions for our users.  

Not only did the Google for Startups Accelerator program help to achieve our business goals for the year, it was also extremely beneficial to be connected with a cohort of other women founders of color. While we are all building different kinds of businesses across different industries, the guidance around fundraising, scaling teams, and coping with the struggles of being a founder rang true for all of us. 

Advancing the community

It can be daunting to launch your own business – and even more so as an AAPI woman. Throughout my career, I am certain I have experienced unconscious bias around my race. Specifically, there have likely been “model minority” stereotypes about my demeanor and math abilities, assumptions that I may be mild-mannered or agreeable. While frustrating, these unfair tropes actually inspired me to see myself beyond others’ perceptions. Why be either analytical or creative, meek or brash, inspiring or agreeable? It’s the ors of life that prevent us from seeing ourselves as truly multi-dimensional individuals. I aspired to be more instead of or, and want my company to be as well. 

As an inclusion-focused business, it is extremely important that Fairygodboss mindfully engage with all underrepresented groups. Diversity can be sliced in many different ways, and these intersections lead to great opportunities for change. We have a team of volunteer ‘culture wizards’ at work that provide educational resources and videos for those wanting to learn more about different cultures and identities from members of under-represented groups. We also practice what we preach at Fairygodboss by prioritizing diverse slates and diversity in our own hiring process. We aim for the talent we hire to be representative of the diverse communities we support. It is imperative to take an intentional approach to hiring diverse talent, as it falls on all employers to promote equity and inclusion—there are always ways in which we can all improve. Fairygodboss looks forward to evolving with Google Cloud in 2021 and beyond as we work together to foster inclusive teams around the world. 

If you want to learn more about how Google Cloud can help your startup, visit our Startup Program application page here and sign up for our monthly startup newsletter to get a peek at our community activities, digital events, special offers, and more.