Hi Jody,
not sure how to identify if the "fast-path" was taken or not, I don't see
anything suggesting it did, but I could be looking for wrong key words.
Thank you for explaining the components. What confused me was that I saw
the image chain and reprojection information logged in but didn't realize
it's just the course of action to be taken later.
I had a look at image processing chains of both requests, JAI operations
used are probably as expected:
1) Native: ImageRead -> Crop -> Scale -> RasterClassifier -> Mosaic
2) Reprojected: ImageRead -> Crop -> Warp -> Crop -> RasterClassifier ->
Mosaic
Here is the full image processing chain when requesting reprojected layer -
any idea why could the additional warp operation slow down the request so
much?
2022-10-20 13:35:43,589 DEBUG [org.geoserver.wms] - setting up map
2022-10-20 13:35:43,590 DEBUG [org.geoserver.wms] - setting up 1385x912
image
2022-10-20 13:35:43,642 DEBUG [org.geoserver.wms.map] - Direct rendering
path produced the following image chain:
JAI op: Mosaic(it.geosolutions.jaiext.mosaic.MosaicOpImage) at Level: 0,
offset:0, 0, size:1385 x 912, tile size:895 x 407
Params. Parameter 1:MOSAIC_TYPE_OVERLAY; Parameter 2:null; Parameter
3:[javax.media.jai.ROI@2b4e135f]; Parameter 4:[[0.0]]; Parameter 5:[1.0];
Parameter 6:[RangeDouble[15.0, 15.0]];
Bands: 1, type: Byte; Color model:class java.awt.image.IndexColorModel,
transparency: Bitmask
Tile cache: it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler: com.sun.media.jai.util.SunTileScheduler@7482407<global>,
parallelism 7, priority 5
Number of sources: 1
JAI op:
RasterClassifier(it.geosolutions.jaiext.classifier.RasterClassifierOpImage)
at Level: 1, offset:-11, -11, size:1407 x 934, tile size:895 x 407
Params. Parameter 1:[Domain description:
name= Label
input range=RangeDouble[1.0, 1.0]
output range=RangeDouble[0.0, 0.0]
colors=java.awt.Color[r=62,g=192,b=224]]; Parameter 2:-1; Parameter
3:javax.media.jai.ROI@c91b673; Parameter 4:RangeDouble[15.0, 15.0];
Bands: 1, type: Byte; Color model:class java.awt.image.IndexColorModel,
transparency: Bitmask
Tile cache:
it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler: com.sun.media.jai.util.SunTileScheduler@7482407<global>,
parallelism 7, priority 5
Number of sources: 1
JAI op: Crop(it.geosolutions.jaiext.mosaic.MosaicOpImage) at Level:
2, offset:-11, -11, size:1407 x 934, tile size:895 x 407
Params. Parameter 1:-11.0; Parameter 2:-11.0; Parameter 3:1407.0;
Parameter 4:934.0; Parameter 5:null; Parameter 6:RangeDouble[15.0, 15.0];
Parameter 7:[15.0];
Bands: 1, type: Byte; Color model:class
java.awt.image.ComponentColorModel, transparency: Opaque
Tile cache:
it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler: com.sun.media.jai.util.SunTileScheduler@7482407<global>,
parallelism 7, priority 5
Number of sources: 1
JAI op: Warp(it.geosolutions.jaiext.warp.WarpNearestOpImage) at
Level: 3, offset:-12, -1313, size:1514 x 3516, tile size:895 x 407
Params. Parameter
1:org.geotools.referencing.operation.transform.WarpAdapter@1b6a660;
Parameter 2:InterpolationNearest; Parameter 3:[15.0]; Parameter 4:null;
Parameter 5:RangeDouble[15.0, 15.0];
Bands: 1, type: Byte; Color model:class
java.awt.image.ComponentColorModel, transparency: Opaque
Tile cache:
it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler:
com.sun.media.jai.util.SunTileScheduler@7482407<global>,
parallelism 7, priority 5
Number of sources: 1
JAI op: Crop(it.geosolutions.jaiext.mosaic.MosaicOpImage) at
Level: 4, offset:0, 33, size:1628 x 895, tile size:512 x 512
Params. Parameter 1:0.0; Parameter 2:33.0; Parameter 3:1628.0;
Parameter 4:895.0; Parameter 5:null; Parameter 6:RangeDouble[15.0, 15.0];
Parameter 7:[15.0];
Bands: 1, type: Byte; Color model:class
java.awt.image.ComponentColorModel, transparency: Opaque
Tile cache:
it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler:
com.sun.media.jai.util.SunTileScheduler@7482407<global>,
parallelism 7, priority 5
Number of sources: 1
JAI op:
ImageRead(com.sun.media.jai.imageioimpl.ImageReadOpImage) at Level: 5,
offset:0, 0, size:1628 x 977, tile size:512 x 512
Params. Parameter 1:FileImageInputStreamExtImpl which points
to PathToFile\RasterFileName.tif.ovr; Parameter 2:5; Parameter 3:false;
Parameter 4:false; Parameter 5:false; Parameter 6:null; Parameter 7:null;
Parameter 8:[class javax.imageio.ImageReadParam SourceSubsampling[ssx:1,
ssy:1] DestinationOffset(Point)[x:0, y:0]]; Parameter
9:it.geosolutions.imageioimpl.plugins.tiff.TIFFImageReader@7fbf999a;
Bands: 1, type: Byte; Color model:class
java.awt.image.ComponentColorModel, transparency: Opaque
Tile cache:
it.geosolutions.concurrent.ConcurrentTileCacheMultiMap@411b2638
Tile scheduler:
com.sun.media.jai.util.SunTileScheduler@7482407<global>, parallelism 7,
priority 5
Number of sources: 0
Thank you for your help,
Paul
st 19. 10. 2022 v 18:11 odesÃlatel Jody Garnett <jody.garn...@gmail.com>
napsal:
> Components:
> - GeoServer uses the geotools library for the heavy lifting (gis data
> formats, referencing, data structures)
> - image processing provided by the java advanced imaging library; with
> JAI-EXT providing GIS operations (to respect concepts like no-data)
> - image formats provided by image-is library, with imageio-ext for GIS
> formats
>
> A goal of the image processing is to be lazy and only do working when
> output is required ... for example when writing content out.
>
> So the native and reproject requests are taking the same time to setup;
> but when starting to write the result out the data will be pulled through
> an image processing chain - and the work required for reprojecting is much
> greater.
>
> There is a fast-path that engages when output and input line up; allowing
> the output to be written out directly from sampling the input. In other
> cases (say when requesting several layers as a single getmap everything
> must be drawn; and then the result encoded into an output image.
>
> I hope this helps - your developer logging may indicate if the "fast-path"
> was taken?
>
> Jody
>
> On Wed, Oct 19, 2022 at 2:45 AM Pa Pu <ppprim...@gmail.com> wrote:
>
>> Hi all,
>>
>> I posted this issue on stackoverflow
>> <https://stackoverflow.com/questions/74114071/geoserver-bottleneck-when-writing-response-image>
>> but it was recommended to me to ask rather here:
>>
>> I've come across a bottleneck that occurs when writing the final png file
>> from a WMS request. I'm rendering a geotiff (tiled, LZW compressed, with
>> external overviews, 90 MB file), reprojecting the layer from native CRS to
>> EPSG:4326. When requesting an image of the same size in native CRS, the
>> request takes around 100 ms, when reprojected, the request takes on average
>> around 7000 ms.
>>
>> I had a look in logs with GEOSERVER_DEVELOPER_LOGGING options and both
>> requests (native/reprojected) are taking similar amount of time except when
>> writing the final png output image, where the request for reprojected image
>> gets stucked:
>>
>> 2022-10-18 14:12:02,059 DEBUG [org.geoserver.wms.map] - Writing png image
>> ...
>> 2022-10-18 14:12:09,139 DEBUG [org.geoserver.wms.map] - Writing png image
>> ... done!
>>
>> Using these log messages, I was able to look up following function in
>> java:
>>
>> public void formatImageOutputStream(
>> RenderedImage image, OutputStream outStream, WMSMapContent
>> mapContent)
>>
>> As I'm not sure what this function does, could anyone advise me where to
>> look next? The server is a virtual machine with Windows Server running in
>> HyperV, multiple cores and plenty of memory allocated to java. The issue
>> seems to be affecting all layers, however same rasters and layers are
>> working fine in other Geoserver instances that I run on physical machines.
>> I understand this might not be a Geoserver issue but I would still love to
>> hear your suggestions on where I could investigate.
>>
>> Thanks a lot.
>>
>> Geoserver version: 2.20.2 Container: Tomcat 9.0 (jre1.8.0_321)
>> _______________________________________________
>> Geoserver-users mailing list
>>
>> Please make sure you read the following two resources before posting to
>> this list:
>> - Earning your support instead of buying it, but Ian Turton:
>> http://www.ianturton.com/talks/foss4g.html#/
>> - The GeoServer user list posting guidelines:
>> http://geoserver.org/comm/userlist-guidelines.html
>>
>> If you want to request a feature or an improvement, also see this:
>> https://github.com/geoserver/geoserver/wiki/Successfully-requesting-and-integrating-new-features-and-improvements-in-GeoServer
>>
>>
>> Geoserver-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/geoserver-users
>>
> --
> --
> Jody Garnett
>
_______________________________________________
Geoserver-users mailing list
Please make sure you read the following two resources before posting to this
list:
- Earning your support instead of buying it, but Ian Turton:
http://www.ianturton.com/talks/foss4g.html#/
- The GeoServer user list posting guidelines:
http://geoserver.org/comm/userlist-guidelines.html
If you want to request a feature or an improvement, also see this:
https://github.com/geoserver/geoserver/wiki/Successfully-requesting-and-integrating-new-features-and-improvements-in-GeoServer
Geoserver-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-users
