[OSM-dev] Tiles at home - current disk use

Frederik Ramm frederik at remote.org
Sat May 19 21:25:18 BST 2007


Hi,

Reply-to set to tiles at home list.

> I've asked a couple of people to try requesting re-renders of everything.  
> Basically, look for tiles generated before 2007-05-19 and request them.

Now that we have a tiles at home mailing list, it would be a good idea to 
communicate such things there, and ask the people you have talked to 
about re-rendering the whole world to post their plans there, lest 
everything is done multiple times.

I'll stop running any update scripts as long as I don't know who these 
people are and what they are planning to do, so that I don't get in 
their way. The stupidest thing to do in the current situation would be 
unnecessarily requesting/uploading tiles multiple times, and avoiding 
that requires communication.

> 5: We could copy images from the old disk to the new. This is probably the 
> worst idea because (a) there's a lot of them, (b) many of them are outdated, 
> and (c) they contain a lot of blank-sea images which we want to remove.

I don't really understand why we'd suddenly treat the disk with the 
blank sea images as "tainted". What's so uncorrectably bad about them? 
Can you not just run

find . -type f -size -600c

and remove the resulting tiles from metadata and filesystem? If you 
remove a few too many or if there are a few sea tiles you don't delete 
with this, that's not too bad since it will fix itself over time 
(when/if these tiles are re-rendered).

If necessary, stop the server for a while while doing this, but it 
shouldn't be terribly hard on the machine. Afterwards, copy the 
remaining tiles to the new disk, or distribute them as you like.

> 7: any other suggestions?

Have you read the (numerous) mailing list posts about the idea of 
seeding the "sea tiles" table with the knowledge we have from Martijn's 
sea tile index? What do you think about this?

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'




More information about the dev mailing list