Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-23 Thread Greg Troxel
Morten Kjeldgaard writes: > Success! I had to set maxnodes to 500K in the splitter and use 4G ram > in the VM to make it work. > > However the resulting gmapsupp.img makes my Legend Cx crash when > it's turned off: it continously beeps and the only way to stop it is > to remove the batteries

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-23 Thread Morten Kjeldgaard
On 22/08/2009, at 12.51, Greg Troxel wrote: So, set the heap size to a bit less than physram and leave it there (e.g. 1700k for 2M physical) to avoid paging use the splitter max-nodes 160 might be a good value, and if you have trouble report it. I have had people say that doesn'

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Apollinaris Schoell
>> > >max-nodes 160 might be a good value, and if you have trouble >report it. I have had people say that doesn't work with --route > and >they need to use 80, but 1.6M works for me. > had to lower it to 1.1M most tiles are much smaller even with 1.6M. splitter doesn't creat

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Johann Gail
> The reason for the high memory usage is that most nodes in the Danish > extract look like this: > > > > > > > > > > > This gets very messy, but I suppose tags could be trimmed on reading if > there is some way to know/declare that they won't be used. The problem >

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Greg Troxel
The reason for the high memory usage is that most nodes in the Danish extract look like this: This gets very messy, but I suppose tags could be trimmed on reading if there is some way to know/declare that they won't be used. The problem is then the relation matching

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Steve Ratcliffe
Hi > I will follow up when I have investigated the large memory usage of > the Danish tiles. Anyone else have problems? The reason for the high memory usage is that most nodes in the Danish extract look like this: All those tags take up the memory. Before relation suppor

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Greg Troxel
Morten Kjeldgaard writes: > On 22/08/2009, at 01.53, Greg Troxel wrote: > >> With splitter/mkgmap, the general strategy I use is: >> >> set heap size to a bit less than RAM, to avoid paging >> >> use --max-nodes to splitter to make small enough areas. >> >> >> Here is the script I use (with a

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-22 Thread Steve Ratcliffe
Hi On 08/22/2009 02:00 AM, Morten Kjeldgaard wrote: > I've tried at 512M, 1024M, all the way up to 4096M, with the same > result. The only difference is at 4G it takes much longer before > mkgmap gives up. I don't think there's much point in going further > than 4G, the problem must lie elsewhere.

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Morten Kjeldgaard
On 21/08/2009, at 23.26, Steve Ratcliffe wrote: > When support for relations was added the memory required increased. I > think that was added along with routing, I believe, which was in > version > 684 or thereabouts, so that will explain what you see. > > I don't use debian, but I just downlo

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Morten Kjeldgaard
On 22/08/2009, at 01.53, Greg Troxel wrote: > With splitter/mkgmap, the general strategy I use is: > > set heap size to a bit less than RAM, to avoid paging > > use --max-nodes to splitter to make small enough areas. > > > Here is the script I use (with a mac that has 4G physical ram): Great! T

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Greg Troxel
Morten Kjeldgaard writes: > On 21/08/2009, at 23.26, Steve Ratcliffe wrote: > >> When support for relations was added the memory required increased. I >> think that was added along with routing, I believe, which was in >> version >> 684 or thereabouts, so that will explain what you see. >> >>

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Morten Kjeldgaard
On 21/08/2009, at 23.26, Steve Ratcliffe wrote: > When support for relations was added the memory required increased. I > think that was added along with routing, I believe, which was in > version > 684 or thereabouts, so that will explain what you see. > > I don't use debian, but I just downlo

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Steve Ratcliffe
On 08/21/2009 09:26 AM, Morten Kjeldgaard wrote: > The default version of mkgmap distributed with Ubuntu 9.04 (jaunty) is > svn630. I tried upgrading to version svn1067 which has recently appeared > in Debian unstable, but now I get a Java heap space exception on the > _same_ .osm file that worke

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Morten Kjeldgaard
On 21/08/2009, at 13.20, Greg Troxel wrote: >> The default version of mkgmap distributed with Ubuntu 9.04 (jaunty) >> is >> svn630. I tried upgrading to version svn1067 which has recently >> appeared in Debian unstable, but now I get a Java heap space >> exception on the _same_ .osm file that

Re: [mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Greg Troxel
Morten Kjeldgaard writes: > The default version of mkgmap distributed with Ubuntu 9.04 (jaunty) is > svn630. I tried upgrading to version svn1067 which has recently > appeared in Debian unstable, but now I get a Java heap space > exception on the _same_ .osm file that worked before. > > The map

[mkgmap-dev] Regression causing Heap space exhaustion?

2009-08-21 Thread Morten Kjeldgaard
Hi, The default version of mkgmap distributed with Ubuntu 9.04 (jaunty) is svn630. I tried upgrading to version svn1067 which has recently appeared in Debian unstable, but now I get a Java heap space exception on the _same_ .osm file that worked before. The map I was working with is the