Changes between Initial Version and Version 1 of MapGuideRfc63


Ignore:
Timestamp:
Apr 21, 2009 7:56:06 PM (10 years ago)
Author:
uvlite
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • MapGuideRfc63

    v1 v1  
     1= !MapGuide RFC 63 - Introducing smart request queues to improve performance and stability =
     2
     3This page contains an change request (RFC) for the !MapGuide Open Source project.
     4More !MapGuide RFCs can be found on the [wiki:MapGuideRfcs RFCs] page.
     5
     6
     7== Status ==
     8
     9||RFC Template Version||(1.0)||
     10||Submission Date||(22.04.2009)||
     11||Last Modified||(UV Wildner) [[Timestamp]]||
     12||Author||(UV Wildner)||
     13||RFC Status||(draft)||
     14||Implementation Status||(discussion needed)||
     15||Proposed Milestone||(2.2?)||
     16||Assigned PSC guide(s)||(when determined)||
     17||'''Voting History'''||(vote date)||
     18||+1||||
     19||+0||||
     20||-0||||
     21||-1||||
     22||no vote||||
     23
     24== Overview ==
     25
     26In the current MapGuide Implementation the multithreaded server receives and processes concurrent requests for computing adjacent tiles.
     27This approach does not make clever use of the machine resources as the parallelization of those operations rarely speeds up the total
     28processing time. Therefore, e propose the introduction of smart request queue which can sort requests by Map & Scale.
     29
     30== Motivation ==
     31
     32A memory limitation can cause the mapping / rendering service to abort rendering the tile before all features are determined.
     33Its very interesting that the serialization enforced by STDOUT of the debug build solves this problem already!
     34Returning the faulty tile without an error message is a defect (like #462)
     35However, this gives us as reason to suggest a redesign of the Tiling service.
     36
     37Basically, the server does not do anything efficient if we start to request mutliple tiles of the same map and scale at the same time.
     38This is like windows copying. If you start n parallel copy streams between the same disks you end up thrashing your disk.
     39
     40== Proposed Solution ==
     41
     421. We therefore propose the introduction of smart request queue into the server which can sort requests by Map & Scale.
     432. The request queue serializes all tile requests and creates separate queues for each mapscale.
     443. A configurable multiplication factor can be set to ask the tileservice to compute larger tiles which are later sliced up into the requested small ones. Using a smart queue there is a way to block requests for those tiles to come in side the queue.
     45
     46This does not seem very hard to do and will improve performance and stability of the mapguide server dramatically.
     47
     48== Implications ==
     49
     50This requires some thoughtful investigation of the queue structures in the ACE. I believe the LoadBalanceManager is a good starting point to look at doing this.
     51
     521. The queue needs to get another dimension based on map and scale information.
     532. Tileservice needs to get wrapped (subclassed) with an implementation [[BR]]
     54
     55    a) which creates large request from the small tiles [[BR]]
     56
     57    b) blocks existing requests for the same map and scale. (maybe even the lockfiles in the tile cache can serve this purpose.[[BR]]
     58
     59    c) splits up the large tile (calling into new methods in the renderers)[[BR]]
     60
     61    d) releases the blocked requests which are requesting tiles computed in the current request (usimg lockfiles in tilecache??)[[BR]]
     62
     63
     64== Test Plan ==
     65
     66TBD
     67
     68== Funding/Resources ==
     69
     70TBD
     71== closed Issues ==
     72
     73NONE
     74
     75== open Issues ==
     76
     77Design
     78
     79
     80== Suggestions ==
     81
     82Please Comment.