Latest Contributions by NathanHeick

I think I had this issue twice before.  I would contact your account manager or customer service. Nathan ... View more

There are other benefits to using the Web Adaptor internally, as well.  For one thing, it can be used to host the ArcGIS Services on port 80 instead of 6080, helping mask the actual server endpoints.  This also lets you restrict access to the administrative REST interface to port 6080.  Most importantly, you can use it to set up Windows Authentication in IIS.  In general, the Web Adaptor acts as a proxy to your actual ArcGIS Server.  I think you still have some more architectural options externally, but I don't know architecture well enough to make any recommendations.  Definitely, consider the Web Adaptor internally, as well. ... View more

I tried creating a topology in ArcSDE 9.3.1 from ArcGIS 10.1 SP1.  I received an error regarding empty geometry.  I then created it directly from ArcGIS 9.3.1.  When I deleted it, I still had errors, but found new messages regarding parcel fabric tables.  When I looked at the SDEINTERCEPT logs, after it deleted this tables it seemed to be going through tables in GDB_ITEMS.  I remember seeing some parcel fabric tables then too.  It's trying to do something after it deletes these topology tables, but I'm not clear what. ... View more

I did some more testing using SDEINTERCEPT.  If I delete the topology using a right-click context menu and review the intercept files, the tables in question do exist, are accessed, deleted, and then additional attempts to access them are made at the end of the process.  Still, there is no apparent failure.  With the GP tool, with Python or in ArcCatalog, something similar happens, but the GP tool fails in response to the Oracle errors.  Is anyone able to build a topology in ArcSDE and ArcGIS 10.2 and then delete it successfully using the Delete tool?  I am wondering if this is a known issue with 10.1.  It feels like a bug to me. ... View more

Hi Val, Try looking at the following URL on paths in Python.  To me, you just need to fix the first two paths.  They use single backslashes.  This looks like you converted a model to Python, but the paths look like they were copied and pasted from ArcCatalog.  When you do that, usually you get single backslashes, the way Windows specifies paths.  You can add an r before the paths, or change the backslashes to double backslashes or single forward slashes.  That will probably fix it.  I paste from ArcCatalog sometimes, and then I get errors until I fix the path strings.  Look up more information about paths in Python.  The problem is that single backslashes are treated like the beginning of an escape sequence, so you either have to use the universal "/" in Python, escape the backslashes ("\\"), or specify that the strings are raw and to read them as is (r""). http://resources.arcgis.com/en/help/main/10.2/index.html#//002z0000000r000000 ... View more

Hello, This looks like the issue of not properly specifying your paths.  Windows uses backslashes and Linux uses forward slashes.  In Python, forward slashes will always work or you can use os.path.sep to get the proper separator for your operating system.  If you use one backslash, Python reads it as the start of an escape code, so you have to escape it with double backslashes.  The final alternative is to use raw strings and add a lower case r before your path like follows.  This reads the string as is, which works correctly in windows. LIVE_CIP_Shapes = r"Database Connections\Connection to bangkok as clvgis.sde\LIVE.CIP_Shapes" IMSV7_CLV_CIPMAP = r"Database Connections\Connection to HANSEN DEV as IMSV7_READ.odc\IMSV7.CLV_CIPMAP" Having run across the same issue before, I imagine you copied and pasted the path from ArcCatalog. Here is a link from the ArcGIS help on paths. http://resources.arcgis.com/en/help/main/10.2/index.html#//002z0000000r000000 ... View more

Responding to the last two questions, the account being used to delete the topology is the data owner.  Also, I had already checked to see if the topology existed before deleting it. Thanks, Nathan ... View more

Actually, after further research, I believe the return value from Pool.apply_async needs to be an object defined at the top-level of any module.  In the end, I decided to write my own Result class that stores the information available through the arcpy.Result methods and lets you retrieve them with an identical interface.  It is picklable if all the inputs and outputs are picklable.  For my case, I'm mostly using datasets represented as path strings. class Result(object):     def __init__(self, status, inputs, outputs, messages):         """Initializes the result.         ARGUMENT            DESCRIPTION         --------            -----------         status              The job status.         inputs              The inputs to the job.         outputs             The outputs of the job.         messages            The messages of the job represented as a list of                             tuples whose values are (severity, message).                 STATUS              DESCRIPTION         ------              -----------         0                   New         1                   Submitted         2                   Waiting         3                   Executing         4                   Succeeded         5                   Failed         6                   Timed Out         7                   Cancelling         8                   Cancelled         9                   Deleting         10                  Deleted                 SEVERITY            DESCRIPTION         --------            -----------         0                   Informational Message         1                   Warning Message         2                   Error Message         """         self._status = status         """The job status.         STATUS              DESCRIPTION         ------              -----------         0                   New         1                   Submitted         2                   Waiting         3                   Executing         4                   Succeeded         5                   Failed         6                   Timed Out         7                   Cancelling         8                   Cancelled         9                   Deleting         10                  Deleted         """         self._inputs = inputs         """The inputs to the job."""         self._outputs = outputs         """The outputs of the job."""         self._messages = messages         """The messages of the job represented as a list of tuples whose values         are (severity, message).                 SEVERITY            DESCRIPTION         --------            -----------         0                   Informational Message         1                   Warning Message         2                   Error Message         """     @classmethod     def from_arcpy_result(cls, result):         """Creates a Result from an arcpy.Result instance.         ARGUMENT            DESCRIPTION         --------            -----------         result              An instance of arcpy.Result.         """         assert isinstance(result, arcpy.Result), (             "Result %r is not an instance of arcpy.Result." % result)         status = result.status         inputs = [result.getInput(i) for i in range(result.inputCount)]         outputs = [result.getOutput(i) for i in range(result.outputCount)]         messages = []         for i in range(result.messageCount):             messages.append((result.getSeverity(i), result.getMessage(i)))         return cls(status, inputs, outputs, messages)     @property     def inputCount(self):         """The number of inputs of the job."""         return len(self._inputs)     @property     def maxSeverity(self):         """The maximum severity of the messages.                 SEVERITY            DESCRIPTION         --------            -----------         0                   Informational Message         1                   Warning Message         2                   Error Message         """         return max([severity for severity, message in self._messages])     @property     def messageCount(self):         """The number of messages."""         return len(self._messages)     @property     def outputCount(self):         """The number of outputs."""         return len(self._outputs)     @property     def resultID(self):         """The job ID if the tool is a geoprocessing service, else ""."""         return self._resultID     @property     def status(self):         """The job status.         STATUS              DESCRIPTION         ------              -----------         0                   New         1                   Submitted         2                   Waiting         3                   Executing         4                   Succeeded         5                   Failed         6                   Timed Out         7                   Cancelling         8                   Cancelled         9                   Deleting         10                  Deleted         """         return self._status     def cancel(self):         """Cancels the job."""         pass     def getInput(self, index):         """Returns the input at the given index, either as a record set or a         string.         ARGUMENT            DESCRIPTION         --------            -----------         index               Index of the input to return.         """         return self._inputs[index]     def getMapImageURL(self, *args, **kwargs):         """Returns a map service image for a given output, if one exists."""         raise NotImplementedError     def getMessage(self, index):         """Returns the message at the given index.         ARGUMENT            DESCRIPTION         --------            -----------         index               Index of the message to return.         """         return self._messages[index]     def getMessages(self, severity=0):         """Returns messages of the given severity or greater.         SEVERITY            DESCRIPTION         --------            -----------         0                   Informational Message         1                   Warning Message         2                   Error Message         """         messages = []         for s, m in self._messages:             if s >= severity:                 messages.append(m)         return "\n".join(messages)     def getOutput(self, index):         """Returns the output at the given index, either as a record set or a         string.         ARGUMENT            DESCRIPTION         --------            -----------         index               Index of the output to return.         """         return self._outputs[index]     def getSeverity(self, index):         """"Returns the severity of the message at the given index.         SEVERITY            DESCRIPTION         --------            -----------         0                   Informational Message         1                   Warning Message         2                   Error Message         """         return self._messages[index][0] ... View more

I am getting the same problem in ArcGIS 10.1 SP1 with the same version of ArcSDE.  If I create a topology and delete it using the Delete tool in a model, I get the same error.  The tables are created by the topology and are deleted by the tool, but an error is still raised. Deleting topology C:\GIS\DatabaseConnections\SDEDEV\arc@SDEDEV.sde\ParcelFabricStaging\pfsCadastralCenterlineTopology... Failed to delete topology C:\GIS\DatabaseConnections\SDEDEV\arc@SDEDEV.sde\ParcelFabricStaging\pfsCadastralCenterlineTopology. Traceback (most recent call last):   File "C:\GIS\ETL\Python\DropTopology.py", line 34, in <module>     arcpy.management.Delete(topology)   File "c:\program files (x86)\arcgis\desktop10.1\arcpy\arcpy\management.py", line 3658, in Delete     raise e ExecuteError: DBMS table not found [DBMS table not found [ORA-00942: table or view does not exist ][ARC.T_1_PolyErrors]] DBMS table not found [DBMS table not found [ORA-00942: table or view does not exist ][ARC.T_1_LineErrors]] DBMS table not found [DBMS table not found [ORA-00942: table or view does not exist ][ARC.T_1_PointErrors]] Failed to execute (Delete). ... View more

After reading more about pickling, it appears that the function result from Pool.apply_async needs to be defined in the top-level of the main module.  Adding the following solved my pickling problem for the results of the geoprocessing tool: from arcpy import Result ... View more

After a lot of review of the multiprocessing module code, I believe the worker process in the process pool is raising an exception when it tries to put the arcpy.Result object on outqueue.  Trying to run the geoprocessing tool manually and pickling the result raised the following PicklingError. PicklingError: Can't pickle 'geoprocessing server result object' object: <geoprocessing server result object object at 0x13DC4698> I guess the moral of the story is that I really need to understand pickling as it relates to arcpy. ... View more

As I'm troubleshooting, I'll keep track of my findings here, in case it helps anyone later on.  I believe that sys.stderr is being redirected to a custom geoprocessing stderr object somewhere in arcpy which I couldn't locate.  Also, line 275 above does not refer to a SystemExit exception, but any other exception.  I suspect I am not giving the arguments for the apply_async call properly.  I think the issue is avoidable, but I don't think the custom stderr object from arcpy has a flush method like the normal stderr. ... View more

When you create the csv.DictWriter, the list of fieldnames determines the order in which the fields are written to the .csv file.  I would recommend modifying your code to write the rows right after you read them in.  With 23 million rows, you don't want to read all of the rows into memory and then write out the data. If you want to keep working on performance, there is a good session from the 2013 DevSummit about geoprocessing in Python and performance.  The newer versions of ArcGIS have a new search cursor that runs significantly faster.  You may have some other alternative approaches that are even faster, possibly using numPy arrays.  Look at the documentation on search cursors.  It's usually good to del row, rows when you are done with the search cursor to release any locks. ... View more