8

I'm suffering from very slow viewing data of and ArcSDE database from an Arcgis engine application, the SDE database is in localhost so there's no network issues. but i still don't know the reasons neither how to solve this problem.

i'm working on 64 bit machine with 2 CPU's Xeon 3.4 GHz and 2 Gb of RAM.

the database contains 20 feature class, but with the max number of features do not exceed 100 000 in some feature classes, if i wait to view the data i should wait 10 minutes.

i tried to compress and rebuild indexes in the database but there's no improvements at all.

i tried to view data from ArcMap, i have founded the same problem.

Using the performance monitor i remarked some bottlenecks in CPU and network sides.

SDEINTERCEPT details:

i have tried @travis advice, so i opened mxd referencing data in sde with Arcmap, then i remarked this part take 7 minutes:

[W 10:34:37.710] Command:      QueryWithInfo
[R 10:34:37.710] Long:         1
[R 10:34:37.710] Query Info: 
    Num Columns:   1
    Columns:       "shape"
    SQL_Construct: [1]
    Tables:        "gebase.sde.point"
    WhereClause:   "type_point_id<3"
    Query Type:    4
    Num Hints:     0
    Num Parameter markers: 0
    Logfile:       <null>
[W 10:34:37.718] Long:         0
[W 10:34:37.718] Col_Defines:  [1]
    Name                                 Type    Width nDec  NULL?   RowID
    -------------------------------- ----------- ----- ---- -------- -----
    shape                            SE_SHAPE        0   0      NULL      
    -------------------------------- ----------- ----- ---- -------- -----
[W 10:34:37.718] Long:         71303299
[W 10:34:37.718] Long:         0
[W 10:34:37.718] CoordRef:
    XY False Origin:       -37644800,, -28128500,
    XY System Units:       10000,
    XY Half SysUnit:       0,00005
    XY Round:              0,0001
    XY Cluster Tolerance:  0,001
    Z  Offset:             -100000,000000
    Z  Units:              10000,000000
    Z  Half SysUnit:       0,000050000
    Z  Round:              0,000100000
    Z  Cluster Tolerance:  0,001
    Measure Offset:        -100000,000000
    Measure Units:         10000,000000
    Measure Half SysUnit:  0,000050000
    Measure Round:         0,000100000
    Measure Cluster Tol:   0,001
    Coordinate System ID:  0
    Coordinate System:     "PROJCS["Nord_Maroc_Degree",GEOGCS["GCS_Merchich_Degree",DATUM["D_Merchich",SPHEROID["Clarke_1880_IGN",6378249.2,293.46602]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic"],PARAMETER["False_Easting",500000.0],PARAMETER["False_Northing",300000.0],PARAMETER["Central_Meridian",-5.4],PARAMETER["Standard_Parallel_1",33.3],PARAMETER["Scale_Factor",0.999625769],PARAMETER["Latitude_Of_Origin",33.3],UNIT["Meter",1.0]]"
    Spatial Reference ID:  102191
    Precision              High [64]
========================================
[W 10:34:37.719] Command:      SetSpatialConstraints
[R 10:34:37.719] Long:         1
[R 10:34:37.719] Short:        2
[R 10:34:37.719] Long:         0
[R 10:34:37.720] Filter Array: [1]
    Table:        gebase.sde.point
    Column:       shape
    SearchMethod: SM_ENVP
    Truth:        Must Pass
    FilterType:   SE_SHAPE_FILTER
          FilterShape:
          XY False Origin:       -37644800,, -28128500,
          XY System Units:       10000,
          XY Half SysUnit:       0,00005
          XY Round:              0,0001
          XY Cluster Tolerance:  0,001
          Z  Offset:             -100000,000000
          Z  Units:              10000,000000
          Z  Half SysUnit:       0,000050000
          Z  Round:              0,000100000
          Z  Cluster Tolerance:  0,001
          Measure Offset:        -100000,000000
          Measure Units:         10000,000000
          Measure Half SysUnit:  0,000050000
          Measure Round:         0,000100000
          Measure Cluster Tol:   0,001
          Coordinate System ID:  0
          Coordinate System:     "PROJCS["Nord_Maroc_Degree",GEOGCS["GCS_Merchich_Degree",DATUM["D_Merchich",SPHEROID["Clarke_1880_IGN",6378249.2,293.46602]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic"],PARAMETER["False_Easting",500000.0],PARAMETER["False_Northing",300000.0],PARAMETER["Central_Meridian",-5.4],PARAMETER["Standard_Parallel_1",33.3],PARAMETER["Scale_Factor",0.999625769],PARAMETER["Latitude_Of_Origin",33.3],UNIT["Meter",1.0]]"
          Spatial Reference ID:  102191
          Precision              High [64]
          Feature Number:        0
          Feature Entity Type:   Area        
          Number of Points:      5
          Feature Envelope:
            MinX:    328133,48150, MaxX:    384094,63650
            MinY:    133834,78230, MaxY:    159869,12210
          Polygon Perimeter:        163990,98960
          Polygon Area:          1456911724,87047
---------------------------------------------------------------
Point          X                Y               2D Distance 
---------------------------------------------------------------
    1     328133,48150     133834,78230
    2     384094,63650     133834,78230        55961,155
    3     384094,63650     159869,12210        26034,340
    4     328133,48150     159869,12210        55961,155
    5     328133,48150     133834,78230        26034,340

[W 10:34:37.721] Long:         0
========================================
[W 10:34:37.721] Command:      ExecuteSpatialQuery
[R 10:34:37.721] Long:         1
[W 10:34:37.727] Long:         0
========================================
[W 10:41:17.554] Command:      NextBuffer
[R 10:41:17.554] Long:         1
[W 10:41:17.554] Long:         0
[R 10:41:17.554] Long:         16416
[W 10:41:17.554] Long:         456
[W 10:41:17.554] Short:        -1
[W 10:41:17.554] Long:         0
[W 10:41:17.554] Long:         0
[W 10:41:17.554] Block:
    BufferInfo: [25/16416]  Address@0x26fb0000 
    BufferInHex:    "02008A850100010000000100140000000C0000000100000082..."

what could make ExecuteSpatialQuery so slow to take 7 minutes ?

I'm available for explanations of the problem.

Any help please.

Improve this question
15
  • 1
    I suspect you will be more likely to get a response if you detail your system architecture, editing workflow, data characteristics, and what steps you've taken to diagnose and attempt to remedy the problem already. If you give a vague question you will most likely get a vague answer, or none at all.
    – blah238
    Commented Apr 11, 2013 at 9:48
  • i have made an edit including more details
    – geogeek
    Commented Apr 11, 2013 at 14:35
  • 2
    What do you mean "compared to ArcMap it must run very fast"? The times you give do sound slow. If the database server has other stuff running on it, you could have an issue there. Usually, the start edit/stop edit would not be called that often -- normally you would say use an insert cursor in between them (or update or delete, etc). I believe featureclass.create() is not a super fast call. If you are refreshing 20 layers with labeling, the refresh doesn't sound completely unreasonable.
    – awesomo
    Commented Apr 11, 2013 at 16:21
  • If you are trying to see several featureclasses at once, some of which contain 100 000 records, of course you will take some time. 10 seconds is very reasonable for this quantity of data.
    – Devdatta Tengshe
    Commented Jun 11, 2013 at 10:42
  • 1
    If you want to know exactly what operation is taking the time, you can try running a DBMS trace or SDE intercept (support.esri.com/en/knowledgebase/techarticles/detail/35704). Once you know what is taking all the time, you may be able to get specific help to resolve it.
    – travis
    Commented Jun 12, 2013 at 0:54

2 Answers 2

Reset to default
6
+50

I'm guessing you may have general performance issues such as labeling, layer scale rendering, transparencies..etc. If you have ArcGIS for Server license you may try using the Analyze button on the Service Editor to test a map document for these performance warnings.

Service Editor helps you identify potential performance bottlenecks and errors that you will need to address

Analyzing your GIS resource

You may then model your fixed map document layer configuration to use within your ArcEngine app to improve performance.

Improve this answer
0

i have found the cause of very slow ArcSDE performance, it seems to be that the Database is aging, due to several Arcpad check-out check-in operations and some export workspace to XML ,so this is not due to Big Data.

So i have created a new geodatabase and imported the XML_workspace and it runs fast compared to old database. so deleting all the feature classes and importing XML_workspace seems not to refresh the database.

is there any way to refresh the database ?

apart the dirty way which is : DROP database and create new geodatabase then import XML_workspace.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.