COMコンポーネントの呼び出しからHRESULT E_FAILエラーが返されました

Question

<\/HEAD>こんにちは、<\/P> 私は研究者で、形態計測解析のためにarcHydroを使用しています。2〜3回の成功した試行の後、正常に動作しなくなりました。DEMの流域グリッド区分後に流域ポリゴン処理ができません。毎回「Hresult e_fail has been returned from a call to a com component」というエラーメッセージが表示されます。以前の試行に関連するすべての一時ファイルを削除し、arcHydroを再インストールしましたが、問題は解決しませんでした。<\/P> このエラーの原因が何であるか、またこの問題の解決策をご存知の方はいらっしゃいますか？<\/SPAN><\/P><\/H1><\/BODY><\/HTML>

BrendenMclane · Answer

In addition to moving my data to my C drive, I found that I also needed to move the Python script, through which I was running ArcHydro, to the folder where my data was stored on my C drive, to avoid the "hresult" error.

Previously the script was on my organization's network and I would progressively lose ArcHydro tools to the "hresult" error. What is very strange is that everything worked on the network at the start of my project (including data storage), but as I did more runs and fleshed out my script, I began encountering the error in one ArcHydro tool after the next. After finding a work around for the tool I needed, the script would run ok for a while and then another tool would break. It was the same whether I was using the default ArcHydro toolbox or the Python toolbox. Finally I had to move everything to local storage and am now encountering no errors.

I am running ArcGIS and ArcHydro 10.5 in IDLE.

Thanks for the tips about local storage.

Brenden

MarkBoucher · Answer

This is a common error,  the solution to which eludes many. The answer could be found in ArcHydro Problem Solvers. Go there and search the page for "hresult" and you'll find some discussion on it.

One of the tips is to run your project on your local PC and not on the network. That is, put the mxd and all geodatabase the project creates on your local machine. There is something about saving the output over a network that slows things down and causes problems.

Also, put your own data in a geodatabase (agree stream, inner walls, etc.). This is more stable than shape files and for some reason I stopped having errors once I did this.

Best,

Mark