[mapserver-users] conf of MapServ+Apache.Lighttpd for better usability

Wim Vanbelle wimvanbelle at gmail.com
Wed Nov 3 07:07:15 EDT 2010


This would be a good thing. That way, if you would zoom in to a location
very quickly, you would also not have to wait for mapserver to render the
outer zooms first (where  you clicked through quite fast), but you would get
your data first and the data you don't need anymore later.

But i suppose this is not easy to realize.

On 2 November 2010 18:40, Sebastian E. Ovide <sebastian.ovide at gmail.com>wrote:

> Hi Guys,
>
> I am using MapServer as a FastCgi (with lighttpd and Apache) connected to
> Oracle, PG, ESRI shapes, ESRI binary grids and MapInfo TAB files.
>
> There is a situation that makes the use of MapServer not very user
> friendly:
>
> An user A opens a map, activates a slow layer and drags the map around.
> This would add a lot of calls to the webserver that would launch many
> MapServer processes reaching the maximum number of processes. At the same
> time, there is another user B displaying another layer... this time a fast
> layer.... As the maximum number of processes has been reached, the webserver
> cannot create new processes until those alive finish their work, it cannot
> serve the other user... that is very annoying as The other user would have
> to wait a long time until the queue is processed and his map is displayed.
>
> Another similar situation is when the user A get annoyed as the map is not
> displaying and therefore decides to turn off the layer and turn on a
> different layer... say a fast layer... he still need to wait that all the
> MapServer processes finish the processing... That annoys the users that
> doesn't understand what is going on... and thinks that the server has
> crashed...
>
> Is there anyway to configure Apache/Lighttpd to give higher priority to the
> new calls ? (that means to conver the queue from FIFO to LIFO....)
>
> I have configured Apache FastCgi to kill the old processes once a minute...
> but the process wont die... actually they would not die with a simple
> kill... they would die with only kill -9 !!!
>
> Any other solutions ?
>
> Thanks
> --
> Sebastian E. Ovide
>
>
>
>
>
> _______________________________________________
> mapserver-users mailing list
> mapserver-users at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/mapserver-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/mapserver-users/attachments/20101103/477b6265/attachment-0001.html


More information about the mapserver-users mailing list