Chyba HRESULT E_FAIL byla vrácena z buňky do komponenty COM v HEC GEO HMS

FauzanAldi

Chyba HRESULT E_FAIL byla vrácena z buňky do komponenty COM, když jsem prováděl zpracování polygonu povodí v hec geo hms. Následoval jsem všechny problémy a návrhy v této komunitě, ale stále to nefunguje a končí to selháním. Prosím o pomoc, přátelé v této komunitě, pomozte vyřešit tento problém.<\/P>

MarkBoucher

@FauzanAldi  - This is a very common error. I think it means that the process can't find the data it is looking for. This is because:

• the data connection is broken - possibly the latent time between the server and your PC is too long and something times out - this is my guess as to why it works with fewer errors on your local PC compare to over the network to a server - faster on a local drive anyway,
• a data or filename is too long for one of the processes - some older ArcHydro routines have legacy file/path/layer name length restrictions - a reason to use the local drive an a simple directory name- raster names have restrictions,
• you missed a process in the sequence - check that you are not missing a key step
• you or some process deleted some intermediate data - in retrying processes this can happen - I sometimes delete all the intermediate data and start again with a clean slate.
• the processing extent has been changed for some reason and the data you want to use exists but is not in the processing extent.

I have struggled to understand the root cause of this error in the ArcHydro process, and the above are some "theories" I try to work with in debugging the error. I've actually made a Model Builder model of a key process using logic and Spatial Analyst tools (a workaround). I've had success with this, but it is time consuming to do this "programing" and using it on the next project requires a lot of checking to make sure file paths are correct or I'll overwrite data in the last project. Learning python better would help in this, I'm sure.

I have not had the chance to use ArcHydro for many months, so I won't offer to take your data and work with it. Possibly someone else has time to do that.

I hope this helps.

Best,

Mark