MapFile.Writer throwing 'File exists error'.

• Key: NUTCH-94
• Type: Bug
• Status: Closed
• Resolution: Duplicate
• Priority: Major
• Assignee: Piotr Kosiorowski
• Reporter: Michael Couck
• Component/s: fetcher
• Affects Version/s: 0.6
• Fix Version/s: 0.7.2

Description

Running Nutch inside a server JVM or multiple times in the same JVM, MapFile.Writer doesn't get collected or closed by the WebDBWriter and the associated files and directories are not deleted, consequently throws a File exists error in the constructor of MapFile.Writer.

Seems that this portion of code is very heavily integrated into Nutch and I am hesitant to look for a solution personally as a retrofit will be necessary with every release.

Has anyone got any ideas, had the same issue, any solutions?

Regards

Michael

Comments

[raghavendra prabhu]

Ya Even I come across the same error

The problem seems to be in the MapFile Writer where it tries to create a newWeb Db and the the directory already exists

I tried to remove the file but is being used by some other resource .So it is not gettign deleted

MapFile.Writer newDb = (comparator == null) ? new MapFile.Writer(fs, newDbFile.getPath(), keyClass, valueClass) : new MapFile.Writer(fs, newDbFile.getPath(), comparator, valueClass);

I tried to delete the newDbFile.getPath()
File dbfileexists=new File(newDbFile.getPath());
if (dbfileexists.exists())
{
System.out.println("already exists");
if(FileUtil.fullyDelete(dbfileexists))

{ System.out.println("deleted successfully"); }

else

{ System.out.println("not deleted"); }

}

and found that it already exists and is not getting deleted when you get the error

[Michael Couck]

Finally it seems that the solution would require too much re-writing and maintaince due to the integration, so we have implemented a spider.

Good luck

Michael

[Piotr Kosiorowski]

Duplicate of NUTCH-117.