Google requires large computational resources in order to provide their services. This article describes the technological infrastructure behind Google's websites, as presented in the company's public announcements.
Original hardware 
- Sun Ultra II with dual 200 MHz processors, and 256 MB of RAM. This was the main machine for the original Backrub system.
- 2 × 300 MHz Dual Pentium II Servers donated by Intel, they included 512 MB of RAM and 10 × 9 GB hard drives between the two. It was on these that the main search ran.
- F50 IBM RS/6000 donated by IBM, included 4 processors, 512 MB of memory and 8 × 9 GB hard drives.
- Two additional boxes included 3 × 9 GB hard drives and 6 x 4 GB hard drives respectively (the original storage for Backrub). These were attached to the Sun Ultra II.
- IBM disk expansion box with another 8 × 9 GB hard drives donated by IBM.
- Homemade disk box which contained 10 × 9 GB SCSI hard drive.
Current hardware 
|This article is outdated. (November 2010)|
Servers are commodity-class x86 PCs running customized versions of Linux. The goal is to purchase CPU generations that offer the best performance per dollar, not absolute performance. How this is measured is unclear, but it is likely to incorporate running costs of the entire server, and CPU power consumption could be a significant factor. Servers as of 2009–2010 consisted of custom made open top systems containing two processors (each with 2 cores), a considerable amount of RAM spread over 8 DIMM slots housing double height DIMMS, and two SATA hard drives connected through a non-standard ATX sized power supply. According to CNET and to a book by Hennessy, each server has a novel 12 volt battery to reduce costs and improve power efficiency.
According to Google their global data center operation electrical power ranges between 500 and 681 megawatts. The combined processing power of these servers might reach from 20 to 100 petaflops.
Network topology 
Details of the Google world wide private networks are not publicly available but Google publications make references to the "Atlas Top 10" report that ranks Google as the 3rd largest ISP behind Level 3 and Global Crossing.
From this site we can see that the Google network can be accessed from 67 public exchange points and 69 different locations across the world. As of May 2012 Google has 882 Gbit/s of public connectivity (which doesn't account for the private peering agreements that Google has with the largest ISP's). This public network is used to distribute content to Google users as well as to crawl the internet to build its search indexes.
The private side of the network is very secret but recent disclosure from Google indicate that they use custom built high-radix switch-routers (with a capacity of 128 x 10 Gigabit Ethernet port) for the WAN. Running no less than two routers per datacenter (for redundancy) we can conclude that the Google network scales in the terabit per second range (with two fully loaded routers the bi-sectional bandwidth amount to 1,280 Gbit/s).
From a datacenter view the network start at the rack level, where Racks are custom-made and contain 40 to 80 servers (20 to 40 1U servers on either side, while new servers are 2U Rackmount systems. Each rack has a switch). Servers are connected via a 1 Gbit/s Ethernet link to the top of rack switch (TOR). TOR switches are then connected to a gigabit cluster switch using multiple gigabit or ten gigabit uplinks. The cluster switches themselves are interconnected and form the datacenter interconnect fabric (most likely using a dragonfly design rather than a classic butterfly or flattened butterfly layout).
From an operation standpoint when a client computer attempts to connect to Google, several DNS servers resolve www.google.com into multiple IP addresses via Round Robin policy. Furthermore, this acts as the first level of load balancing and directs the client to different Google clusters. A Google cluster has thousands of servers and once the client has connected to the server additional load balancing is done to send the queries to the least loaded web server. This makes Google one of the largest and most complex content delivery networks.
Data centers 
Google has numerous data centers scattered around the world. At least 12 significant Google data center installations are located in the United States. The largest known centers are located in The Dalles, Oregon; Atlanta, Georgia; Reston, Virginia; Lenoir, North Carolina; and Moncks Corner, South Carolina. In Europe, the largest known centers are in Eemshaven and Groningen in the Netherlands and Mons, Belgium. Google's Oceania Data Center is claimed to be located in Sydney, Australia. 
Project 02 
One of the larger Google data centers is located in the town of The Dalles, Oregon, on the Columbia River, approximately 80 miles from Portland. Codenamed "Project 02", the $600 million complex was built in 2006 and is approximately the size of two American football fields, with cooling towers four stories high. The site was chosen to take advantage of inexpensive hydroelectric power, and to tap into the region's large surplus of fiber optic cable, a remnant of the dot-com boom. A blueprint of the site has appeared in print.
Summa papermill 
In February 2009, Stora Enso announced that they had sold the Summa paper mill in Hamina, Finland to Google for 40 million Euros. Google plans to invest 200 million euros on the site to build a data center. Google chose this location due to the availability and proximity of renewable energy sources.
Modular Container Data Centers 
Most of the software stack that Google uses on their servers was developed in-house. It is believed that C++, Java, and Python are favored over other programming languages. For example, the back end of Gmail is written in Java and the back end of Google Search is written in C++. Google has acknowledged that Python has played an important role from the beginning, and that it continues to do so as the system grows and evolves.
The software that runs the Google infrastructure includes:
- Google Web Server (GWS) — Custom Linux-based Web server that Google uses for its online services.
- Storage systems:
- Google File System and its successor, Colossus
- BigTable — structured storage built upon GFS/Colossus
- Spanner — planet-scale structured storage system, next generation of BigTable stack
- Google F1 — a distributed, quasi-SQL DBMS based on Spanner, substituting a custom version of MySQL.
- Chubby lock service
- Borg — job scheduling and monitoring system
- MapReduce and Sawzall programming language
- Indexing/search systems:
Google has developed several abstractions which it uses for storing most of its data:
- Protocol Buffers — "Google's lingua franca for data", a binary serialization format which is widely used within the company.
- SSTable (Sorted Strings Table) — a persistent, ordered, immutable map from keys to values, where both keys and values are arbitrary byte strings. It is also used as one of the building blocks of BigTable.
- RecordIO — a sequence of variable sized records.
Software development practices 
Most operations are read-only. When an update is required, queries are redirected to other servers, so as to simplify consistency issues. Queries are divided into sub-queries, where those sub-queries may be sent to different ducts in parallel, thus reducing the latency time.
Search infrastructure 
Like most search engines, Google indexes documents by building a data structure known as inverted index. Such an index allows obtaining a list of documents by a query word. The index is very large due to the number of documents stored in the servers.
The index is partitioned by document IDs into many pieces called shards. Each shard is replicated onto multiple servers. Initially, the index was being served from hard disk drives, as is done in traditional information retrieval (IR) systems. Google dealt with the increasing query volume by increasing number of replicas of each shard and thus increasing number of servers. Soon they found that they had enough servers to keep a copy of the whole index in main memory (although with low replication or no replication at all), and in early 2001 Google switched to an in-memory index system. This switch "radically changed many design parameters" of their search system, and allowed for a significant increase in throughput and a large decrease in latency of queries.
In June 2010, Google rolled out a next-generation indexing and serving system called "Caffeine" which can continuously crawl and update the search index. Previously, Google updated its search index in batches using a series of MapReduce jobs. The index was separated into several layers, some of which were updated faster than the others, and the main layer wouldn't be updated for as long as two weeks. With Caffeine the entire index is updated incrementally on a continuous basis. Later Google revealed a distributed data processing system called "Percolator" which is said to be the basis of Caffeine indexing system.
Server types 
- Google web servers coordinate the execution of queries sent by users, then format the result into an HTML page. The execution consists of sending queries to index servers, merging the results, computing their rank, retrieving a summary for each hit (using the document server), asking for suggestions from the spelling servers, and finally getting a list of advertisements from the ad server.
- Data-gathering servers are permanently dedicated to spidering the Web. Google's web crawler is known as GoogleBot. They update the index and document databases and apply Google's algorithms to assign ranks to pages.
- Each index server contains a set of index shards. They return a list of document IDs ("docid"), such that documents corresponding to a certain docid contain the query word. These servers need less disk space, but suffer the greatest CPU workload.
- Document servers store documents. Each document is stored on dozens of document servers. When performing a search, a document server returns a summary for the document based on query words. They can also fetch the complete document when asked. These servers need more disk space.
- Ad servers manage advertisements offered by services like AdWords and AdSense.
- Spelling servers make suggestions about the spelling of queries.
- "Google Stanford Hardware." Stanford University (provided by Internet Archive). Retrieved on July 10, 2006.
- Tawfik Jelassi and Albrecht Enders (2004). "Case study 16 — Google". Strategies for E-business. Pearson Education. p. 424. ISBN 978-0-273-68840-2.
- Computer Architecture, Fifth Edition: A Quantitative Approach, ISBN 978-0123838728; Chapter Six; 6.7 "A Google Warehouse-Scale Computer" page 471 "Designing motherboards that only need a single 12-volt supply so that the UPS function could be supplied by standard batteries associated with each server"
- Google's secret power supplies
- Google on-server 12V UPS, 1 April 2009.
- Google Green infographics
- Analytics Press Growth in data center electricity use 2005 to 2010
- Google Surpasses Supercomputer Community, Unnoticed?, May 20, 2008.
- "Fiber Optic Communication Technologies: What's Needed for Datacenter Network Operations", Research, Google.
- "FTTH look ahead — technologies & architectures", Research, Google.
- James Pearn. How many servers does Google have?. plus.google.com.
- "Google ASN15169", Peering DB.
- "Urs Holzle", Speakers, Open Network Summit.
- Web Search for a Planet: The Google Cluster Architecture (Luiz André Barroso, Jeffrey Dean, Urs Hölzle)
- Warehouse size computers
- Denis Abt High Performance Datacenter Networks: Architectures, Algorithms, and Opportunities
- Fiach Reid (2004). "Case Study: The Google search engine". Network Programming in .NET. Digital Press. pp. 251–253. ISBN 978-1-55558-315-6.
- Google data center landing page
- Google Data center locations
- Rich Miller (March 27, 2008). "Google Data Center FAQ". Data Center Knowledge. Retrieved 2009-03-15.
- Brett Winterford (March 5, 2010). "Found: Google Australia's secret data network". ITNews. Retrieved 2010-03-20.
- Google "The Dalles, Oregon Data Center" Retrieved on January 3, 2011.
- Markoff, John; Hansell, Saul. "Hiding in Plain Sight, Google Seeks More Power." New York Times. June 14, 2006. Retrieved on October 15, 2008.
- Strand, Ginger. "Google Data Center" Harper's Magazine. March 2008. Retrieved on October 15, 2008.[dead link]
- "Stora Enso divests Summa Mill premises in Finland for EUR 40 million". Stora Enso. 2009-02-12. Retrieved 12.02.2009.
- "Stooora yllätys: Google ostaa Summan tehtaan". Kauppalehti (in (Finnish)) (Helsinki). 2009-02-12. Retrieved 2009-02-12.
- "Google investoi 200 miljoonaa euroa Haminaan". Taloussanomat (in (Finnish)) (Helsinki). 2009-02-04. Retrieved 2009-03-15.
- Finland – First Choice for Siting Your Cloud Computing Data Center. Accessed 4 August 2010.
- "United States Patent: 7278273". Patft.uspto.gov. Retrieved 2012-02-17.
- Mark Levene (2005). An Introduction to Search Engines and Web Navigation. Pearson Education. p. 73. ISBN 978-0-321-30677-7.
- "Python Status Update". Artima. 2006-01-10. Retrieved 2012-02-17.
- [http: //panela.blog-city.com/python_at_google_greg_stein__sdforum.htm "Warning"]. Panela. Blog-city. Retrieved 2012-02-17.
- "Quotes about Python". Python. Retrieved 2012-02-17.
- "Google Architecture". High Scalability. 2008-11-22. Retrieved 2012-02-17.
- Fikes, Andrew (July 29, 2010), "Storage Architecture and Challenges" (PDF), TechTalk, Google.
- Dean, Jeffrey 'Jeff' (2009), "Design, Lessons and Advice from Building Large Distributed Systems" (keynote talk presentation), Ladis, Cornell.
- Shute, Jeffrey 'Jeff'; Oancea, Mircea; Ellner, Stephan; Handy, Benjamin 'Ben'; Rollins, Eric; Samwel, Bart; Vingralek, Radek; Whipkey, Chad; Chen, Xin; Jegerlehner, Beat; Littlefield, Kyle; Tong, Phoenix (2012), "F1 — the Fault-Tolerant Distributed RDBMS Supporting Google's Ad Business" (presentation), Research, Sigmod: Google.
- Intel. Seizing the Open Source Cloud Stack Opportunity. See slide "Proprietary Cloud Computing Stacks".
- "Anna Patterson – CrunchBase Profile". Crunchbase.com. Retrieved 2012-02-17.
- The Register. Google Caffeine jolts worldwide search machine
- "Google Developing Caffeine Storage System | TechWeekEurope UK". Eweekeurope.co.uk. 2009-08-18. Retrieved 2012-02-17.
- "Developer Guide – Protocol Buffers – Google Code". Code.google.com. Retrieved 2012-02-17.
- [dead link]
- Posted by windley on June 24, 2008 1:10 PM (2008-06-24). "Phil Windley's Technometria | Velocity 08: Storage at Scale". Windley.com. Retrieved 2012-02-17.
- "Message limit – Protocol Buffers | Google Groups". Groups.google.com. Retrieved 2012-02-17.
- "Jeff Dean's keynote at WSDM 2009" (PDF). Retrieved 2012-02-17.
- Daniel Peng, Frank Dabek. (2010). Large-scale Incremental Processing Using Distributed Transactions and Notifications. Proceedings of the 9th USENIX Symposium on Operating Systems Design and Implementation.
- The Register. Google Percolator – global search jolt sans MapReduce comedown
- Chandler Evans (2008). "Google Platform". Future of Google Earth. Madison Publishing Company. p. 299. ISBN 978-1-4196-8903-1.
- Chris Sherman (2005). "How Google Works". Google Power. McGraw-Hill Professional. pp. 10–11. ISBN 978-0-07-225787-8.
- Michael Miller (2007). "How Google Works". Googlepedia. Pearson Technology Group. pp. 17–18. ISBN 978-0-7897-3639-0.
Further reading 
- L.A. Barroso, J. Dean, and U. Hölzle (March/April 2002). "Web search for a planet: The Google cluster architecture" (PDF). IEEE Micro 23 (2): 22–28. doi:10.1109/MM.2003.1196112.
- Shankland, Stephen, CNET news "Google uncloaks once-secret server." April 1, 2009.
- Google Research Publications
- Web Search for a Planet: The Google Cluster Architecture (Luiz André Barroso, Jeffrey Dean, Urs Hölzle)
- Underneath the Covers at Google: Current Systems and Future Directions (Talk given by Jeff Dean at Google I/O conference in May 2008)