Spatial Databox and related technologies are owned by Kevin Macdonald, Chris Lamoureux and Marc Lamoureux.
For sales-related enquiries, please email us at cik_mac AT yahoo.ca
For general Google Map development enquiries, and Google Map API licenses, we refer you to this well-known vendor of Google Maps solutions.
Software as a Service
SDB is offered as a Software as a Service (or SaaS) where we host and manage your spatial data on behalf of your business applications, map mashups, etc. To use this service, you pay us service fees based on the number of layers of data under management as well as anticipated or actual demand for access to this data.
Layers
A layer is a collection of points of interest (POI) that share a common structure. For example, all rows within a single database table would constitute a single layer within SDB.
A POI can contain one or more tags, that act as metadata to classify or organize the POI within the context of its layer.
You supply us with two CSV formatted text files for each layer: one file contains the layer records, and the second file contains its tags. These text files can use any common character encoding format (eg. ASCII, ISO-8859-1, UTF-8, etc).
Once loaded into SDB, layer data remains static and cannot be modified directly. To update the data, you will need to supply us with updated text files, which we will apply during the next scheduled weekly service restart. If you require more frequent updates please contact us for pricing details.
Service Utilization
When you turn your office lights off in the evening, your cost for electricity reflects this decrease in consumption. The same concept applies with SDB. As SDB is delivered as a service and priced according to actual use or reserved capacity, you pay only for services that directly benefit your user base.
The level of service chosen will be affected on the following factors:
- type of user (is the user affected by the responsiveness of your application?)
- number of and variance in user traffic
- the nature of the queries that your application will submit to SDB
- opportunity to reuse cached results from previous queries (a map-based application generates one query each time the user interacts with the map and / or its controls / filters. The opportunity to use cached data means fewer queries that need to be submitted to SDB.)
We can help you with capacity planning. As a start, the following sections provide a structure by which you can understand capacity planning in the context of selecting the level of SDB service that balances cost with user satisfaction and productivity.
As a start we find it helpful to first identify the type of user you expect to service: back-office or public-facing.
BACK-OFFICE
A back-office application is used by internal staff or business-to-business (B-B) operations. Demand for service is predictable with gradual changes in traffic volumes, throughout the day. Demand peaks are uncommon. Unlike visitors to publicly accessible websites, back-office users are more patient, as they are paid to interact with the application and will tolerate a greater degree of variance in application responsiveness.
For back-office applications, we offer the following reserved capacity pricing structures (per layer of data):
Small service (100 QPS* / Under 10,000 POI)
- $200 / month
- + $5 / gigabyte of data transferred
Medium service (500 QPS* / Under 1,000,000 POI)
- $800 / month
- + $5 / gigabyte of data transferred
High capacity service (2,000 QPS* / Under 2,500,000 POI)
- $1,800 / month
- + $5 / gigabyte of data transferred
* Stated QPS are for select all queries against any subset of the layer; actual performance can be SIGNIFICANTLY REDUCED for more complex queries.
An example of a complex query is one that applies a tag match against a large percentage of the layer's POI. To maintain stated QPS for this type of complex query, we recommend that additional tag layers be created that comprise only those POI that possess the tag to be queried. Each tag layer carries no additional cost, but does count against the total number of POI available.
Example Use Cases
For applications where the user quickly locates the desired POI, and then interacts with another system or agent on behalf of that POI (eg. copies information into another system or interacts with a client over the phone) the ratio of map interaction vs "think time" is very low, and thus we we expect the Medium service to accommodate from 100 - 500 look-up users.
For applications where the desired POI (or cluster of POI) is not know beforehand, and where a wide range of query conditions and tag selections are performed, the ratio of map interaction vs "think time" is relatively high, and thus we expect the Medium service to accommodate from 5 - 25 analytic users.
PUBLIC-FACING WEBSITE
Popular public-facing websites can experience wide fluctuations in traffic, which in-turn can generate wide fluctuations in demand for services from SDB. For example, a website can demand 1 QPS in the middle of the night, 50 QPS in mid-morning, and surge to 1,000 QPS for a brief duration, immediately after being linked to from a lead story in the New York Times. Unexpected surges are a result of the Slashdot Effect and can complicate capacity planning.
To handle wide flucations in traffic and to minimize cost when demand is low, we offer another price structure based on actual usage and not reserved capacity (priced per per layer of data):
- $400 / month (up to 100,000 POI)
- + $5 / gigabyte of data transferred
- + $5 / CPU hour (CPU equal to 1.2 GHz Intel x86 processor)
As mentioned earlier, SDB performance can be greatly affected by the kind of query performed: a query can use as little as 1 / 10,000 second of CPU up to 1 / 10 second of CPU.
NOTE:
- TERMS, DETAILS AND PRICING ON THIS PAGE ARE SUBJECT TO CHANGE.
- PRICING IS IN US DOLLARS.
- PRICES DO NOT INCLUDE SETUP OR OTHER ADMINISTRATION FEES.
- CLIENT PAYS A RETAINER FEE FROM WHICH SERVICE FEES ARE DEDUCTED ON A DAILY BASIS, PRORATED FROM THE MONTHLY FEE.
Last Updated: Wednesday, 29-July-2010