Hi again Oliver,<div><br></div><blockquote class="gmail_quote" style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0.8ex; border-left-width: 1px; border-left-color: rgb(204, 204, 204); border-left-style: solid; padding-left: 1ex; ">
Uh, yeah. It should be possible to manually partition the data with the features of <span class="J-JK9eJ-PJVNOc" style="background-image: initial; background-attachment: initial; background-origin: initial; background-clip: initial; background-color: yellow; background-position: initial initial; background-repeat: initial initial; ">MapProxy</span> if you only need <span class="J-JK9eJ-PJVNOc" style="background-image: initial; background-attachment: initial; background-origin: initial; background-clip: initial; background-color: yellow; background-position: initial initial; background-repeat: initial initial; ">WMS</span> services. But I don't suggest to do that :)</blockquote>
<div><br></div><div>
I know you warned me, but I am lacking time for some answers on SQLite changes... :)</div><div>So I did partition my data manually using WMS source features, by "cutting of" several sources (the same request) by resolution levels. Additionally I have set the equivalent number of caches, also "cutting of" cache properties with the property "<i>use_direct_from_res</i>" to match sources resolutions (some kind of cascading).</div>
<div>Now I have 5DBs SQLite and tested requests for the resolution different levels (11) and everything seems ok. Tiles are saved where they should be. Also seeding task seems ok, despite it seems to take a little longer, as now it needs to seed 5 different caches.</div>
<div><br></div><div>My question is: Do you feel this approach maybe not be advisable as it can have some drawback with a large amount of data (2Tb).</div><div><br></div><div>Cheers!</div><div><br><div class="gmail_quote">
2011/11/8 Oliver Tonnhofer <span dir="ltr"><<a href="mailto:olt@omniscale.de" target="_blank">olt@omniscale.de</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div>
<br>
On 07.11.2011, at 19:56, Fernando Ribeiro wrote:<br>
> On the other hand, of course we don't want the full 2Tb of data in a single database... we are thinking on a db "switch" scenario for some given zoom values and on minor scales by regions, as MapProxy deals great with these features...but this is only a scenario<br>
<br>
</div>Uh, yeah. It should be possible to manually partition the data with the features of MapProxy if you only need WMS services. But I don't suggest to do that :)<br>
<div><br>
<br>
> Another question;<br>
> Is it safe to install your PIL modification package side-by-side with the "official" PIL package? Should I uninstall it first?<br>
<br>
</div>Pip/easy_install should overwrite/update your existing PIL installation. You can run both versions in different virtualenvs if you have to. I would uninstall PIL if you have installed it as a system package from your distribution.<br>
<div><div><br>
Regards,<br>
Oliver<br>
<br>
--<br>
Oliver Tonnhofer | Omniscale GmbH & Co KG | <a href="http://omniscale.de" target="_blank">http://omniscale.de</a><br>
<a href="http://mapproxy.org" target="_blank">http://mapproxy.org</a> | <a href="https://bitbucket.org/olt" target="_blank">https://bitbucket.org/olt</a> | @oltonn<br>
<br>
<br>
<br>
<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div>Fernando Ribeiro</div><br>
</div>