For cartes.app, we calculate planet wide pmtiles every week, on a hourly-billed server with 384 Go RAM. It worked as usual with the planet pbf file of June 8th, but it crashed with the ones of June 15th and June 22th.
We noticed our tilemaker version was out of date, so we updated it to the lastest commit (42d8b8a).
We also increased the RAM to 512 Go.
But we got the same result : it worked with the planet.osm.pbf file from June 8th and it crashed with June 22th.
We use a lua file specific from cartes.app,, derived from your process-openmaptiles.lua but augmented with quite at lot of stuff now. We did not check if the calculation is working with your lua file. Did anyone here make the calculation for planet wide pmtiles after June 15th ?
We suspect that an OSM element was created or modified between June 8th and June 15th, and that its size makes the RAM to crash. Is it a plausible hypothesis ? I noticed in the lua script the exception for Saimaa lake here so I guess it could happen with another element ?
What do you suggest as a log to identify huge elements that may make the RAM to crash ?
Thanks !
Etienne and @laem
For cartes.app, we calculate planet wide pmtiles every week, on a hourly-billed server with 384 Go RAM. It worked as usual with the planet pbf file of June 8th, but it crashed with the ones of June 15th and June 22th.
We noticed our tilemaker version was out of date, so we updated it to the lastest commit (42d8b8a).
We also increased the RAM to 512 Go.
But we got the same result : it worked with the planet.osm.pbf file from June 8th and it crashed with June 22th.
We use a lua file specific from cartes.app,, derived from your
process-openmaptiles.luabut augmented with quite at lot of stuff now. We did not check if the calculation is working with your lua file. Did anyone here make the calculation for planet wide pmtiles after June 15th ?We suspect that an OSM element was created or modified between June 8th and June 15th, and that its size makes the RAM to crash. Is it a plausible hypothesis ? I noticed in the lua script the exception for Saimaa lake here so I guess it could happen with another element ?
What do you suggest as a log to identify huge elements that may make the RAM to crash ?
Thanks !
Etienne and @laem