Drupal Maps, Clustering.
•
2 likes•1,897 views
The document discusses mapping large numbers of points in Drupal views and some solutions to improve performance when displaying many map points, including using the Leaflet and OpenLayers libraries, enabling client-side clustering with modules like Leaflet Marker Cluster, and progressively enhancing server-side functionality with client-side clustering using the Geocluster module. It provides examples of function call counts and speeds with different mapping approaches and libraries for datasets with thousands of records.
Drupal Maps, Clustering.
- 1. Drupal Maps Showing big amount of map points in views
- 2. Typical Drupal Map Views Recipe views leaflet geocoder geofield addressfield openlayers etc.. etc..
- 3. IT WORKS! The End, EASY!
- 4. You want what?! 2000 map points in the map? sure .. Drupal can do that easy. HEHE I just set the limit of views to no limit! Yay no limit views!
- 5. Your map looks ugly?
- 6. Your map page is slow? There's actually a drupal-tific explaination for that: http://drupal.org/node/1828584 It's a troublesome space trying to deal with presenting large amounts of map points on a map. Typically in drupal space because of other implementations of drupal views mapping integration we experience an exponential theme function calls that directly dependent on the number of records and fields on a view. Thus we encounter "slow" pages in drupal views map pages. Hey with a dataset of 2500+ records and 3 fields having an average number of function calls of over a million plus. Yup it's going slow. For example 2500+ records, 3 fields. Openlayers http://drupal.org/project/openlayers : average of 4.1 million function calls Leaflet http://drupal.org/project/leaflet : average of 1.3 million function calls Leaflet with clientside geoclustering http://drupal.org/project/leaflet_markercluster : average of 1.3 million function calls IMPRESSED?! Now lets move on to bill charges..
- 7. What those numbers don't mean much to you? Current MAP Profile (Openlayers) Executed 90 queries in 240.71 ms. Queries exceeding 5 ms are highlighted. Overall Summary Total Incl. Wall Time (microsec): 104,400,257 microsecs Total Incl. CPU (microsecs): 104,150,000 microsecs Total Incl. MemUse (bytes): 183,759,392 bytes Total Incl. PeakMemUse (bytes): 261,900,368 bytes Number of Function Calls: 5,615,141 (Leaflet) Before Client Side Clustering Executed 7831 queries in 1920.3 ms. Queries exceeding 5 ms are highlighted. Total Incl. Wall Time (microsec): 27,077,213 microsecs Total Incl. CPU (microsecs): 26,560,000 microsecs Total Incl. MemUse (bytes): 242,216,840 bytes Total Incl. PeakMemUse (bytes): 242,562,616 bytes Number of Function Calls: 1,303,140 See Screenshot: Screenshot-leaflet-before-geoclustering-clientside.png
- 8. I heard from the interwebs that you must cluster. http://drupal.org/project/leaflet_markercluster
- 9. WHAT?! your map page is slow? There's actually drupal-tific explaination for that. (Leaflet) After Client Side Clustering Executed 7826 queries in 1913.9 ms. Queries exceeding 5 ms are highlighted. Overall Summary Total Incl. Wall Time (microsec): 27,534,700 microsecs Total Incl. CPU (microsecs): 27,030,000 microsecs Total Incl. MemUse (bytes): 236,385,248 bytes Total Incl. PeakMemUse (bytes): 236,872,064 bytes Number of Function Calls: 1,302,893 See Screenshot: Screenshot-leaflet-after-clientside-clustering.png IMPRESSED?! Now lets move on to bill charges..
- 10. OK! I tawked to an consultant She sez best thing evar is http://drupal. org/project/geocluster
- 11. What she clicked how many times? The system admin is reporting too many ajax calls..
- 12. We have a plan..
- 13. Ok.. its close but we need more time ;) Progressively enhance server-side with client-side clustering http://drupal.org/node/1914704 See you at the issue queues. http://www.drupal.org/project/geocluster