New server and fund raising drive 2013

From OpenStreetMap Wiki
Jump to navigation Jump to search
Help us raise money for a new server to go here!

This funding drive took place in 2013 and was wrapped up in August ahead of the SOTM conference.

You can still donate of course!

Old information

With your help we will raise £40,000 (47k EUR, 61k USD) + £32,500! We will be investing in server infrastructure to improve the performance and stability of OpenStreetMap, and to help the project continue to grow and thrive. Our data contributor volunteers should be able to upload their edits faster and more reliably. OpenStreetMap is growing fast. Our millionth user recently signed up, and we have logged over 15 million changesets.

We are asking for your help. Please help by visiting or promoting our fundraising site:

See Donate for alternative methods.

Thank you very much to all people who donated, already! View their comments.

What's the money for?

Primarily, an additional master database server and other key hardware upgrades to improve reliability and performance. The server will be used at the heart of OpenStreetMap - where the geographic data and edits are stored. Currently only a single server performs this important role, meaning that if there are any problems with it, our mappers are unable to edit the map. With our new ‘Add a note’ feature now live, and a user-friendly HTML5 editor recently launched, we need to ensure OpenStreetMap is always ready to accept edits from the new mappers we’re working to attract.

This planned database server upgrade will greatly increase the reliability and performance of OSM's website and editing service, making it faster for experienced and new mappers alike. It is essential work to enable the OSM community to continue producing the most extensive and up-to-date map of the world.

The operations team has put a lot of thought into designing the new server for maximum performance. Together, these investments will lay the foundation for the continuing growth of OpenStreetMap over the years to come.

What does this mean?

The new server is the first step in putting in place a system which provides more reliable and faster access to the OSM website and API. We've had some problems over the past month, and we're now gradually implementing a redesigned infrastructure which is not reliant on a single server to stay running. The additional server will be available to step in if there should be a problem with the primary machine - this is called redundancy. But even as long as the primary server runs smoothly, the new server will not sit idle, but will be used for spreading work across several machines to provide a faster response - this is called scaling.


Additional servers will allow the resolution of problems without the need for the API or website being unavailable, or read-only. Recently, several hardware failures have resulted in the website or API being unavailable and physical access was required to fix it. Many thanks to Grant for his awesome efforts in getting to, and fixing, these problems quickly. The new server is only a further step towards a broader infrastructure where we try to avoid any single point of failure. With such added redundancy, the risk of of hardware failures causing disruption for everyone using OSM will be greatly reduced.


Most of the accesses to the OSM website and API are read-only requests that can be spread across multiple servers. Many requests involve so much data (e.g: relation/full) that we cannot currently allow them the time to complete - they “time out” because we don’t want them to impede editing activity. Scaling out the database means that such requests can be handled independent of editing activity. They complete faster, and everyone using OSM gets their results faster, and sees fewer time-outs.

Technical specs

4U server (72 - 2.5" bays)
2x Intel Xeon 8 core - 2.4 Ghz
2x LSI 9271 RAID Controller
1x LSI 2308 SAS Controller
25x 900GB 2.5" Disk (Database table spaces)
2x 600GB 2.5" (OS)
2x Intel DC S3700 SSD - 400GB (Index, WAL)
3 year warranty + support. - ORDERED + DELIVERED.

Other Pending Hardware Purchases

  • Replacement for horntail. - DONE
  • SSD + Disks for Gorilla-server.svgorm to make primary rendering server. - DONE


We have extended the donation drive by £32500 to bring forward additional planned hardware upgrades. (blog announcement)

  • Routing Servers (
  • Additional Networking Capacity
  • Additional Tile Server Infrastructure.
  • Additional Redundancy
  • Extending hardware contingency fund