duplicate hadoop temp files
---------------------------
Key: NUTCH-829
URL: https://issues.apache.org/jira/browse/NUTCH-829
Project: Nutch
Issue Type: Bug
Components: generator
Affects Versions: 1.0.0, 1.1
Reporter: Mike Baranczak
Priority: Minor
When two crawls are started at exactly the same time, I see the following
error:
{quote}
org.apache.hadoop.mapred.FileAlreadyExistsException: Output directory
file:/tmp/hadoop-mike/mapred/temp/generate-temp-1276463469075 already exists
at
org.apache.hadoop.mapred.FileOutputFormat.checkOutputSpecs(FileOutputFormat.java:111)
at org.apache.hadoop.mapred.JobClient.submitJob(JobClient.java:793)
at org.apache.hadoop.mapred.JobClient.runJob(JobClient.java:1142)
at org.apache.nutch.crawl.Generator.generate(Generator.java:472)
at org.apache.nutch.crawl.Generator.generate(Generator.java:409)
[...]
{quote}
I traced it down to this code in Generator (I'm using Nutch 1.0, but this is
still in the trunk):
{quote}
Path tempDir =
new Path(getConf().get("mapred.temp.dir", ".") +
"/generate-temp-"+ System.currentTimeMillis());
{quote}
I admit that this is an unlikely scenario for most users, but it just so
happens that I ran into it. To absolutely guarantee that the temp directory
doesn't already exist, I suggest changing System.currentTimeMillis() to
java.util.UUID.randomUUID().toString().
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
