Why connection is terminating

2020-07-11 04:40发布

问题:

I'm trying a random forest classification model by using H2O library inside R on a training set having 70 million rows and 25 numeric features.The total file size is 5.6 GB.

The validation file's size is 1 GB.

I have 16 GB RAM and 8 core CPU on my system.

The system successfully able to read both of the files in H2O object.

Then I'm giving below command to build the model:

model <- h2o.randomForest(x = c(1:18,20:25), y = 19, training_frame = traindata, 
                          validation_frame = testdata, ntrees = 150, mtries = 6)

But after few minutes (without generating any tree), I'm getting following error:

"Error in .h2o.doSafeREST(conn = conn, h2oRestApiVersion = h2oRestApiVersion, : Unexpected CURL error: Recv failure: Connection reset by peer"

However If I tried above code with 1 tree, its running successfully.

Is the above error occurring because of memory issue? Any help will be appreciated.

回答1:

Probably you're out of memory. Try looking on system's memory usage during forest growing. Also try to launch training directly from H2O web console (http://localhost:54321/ by default), may be it will give more detailed error.



回答2:

Its an OutOfMemoryError. A variation of this error message on the R side is:

Error in .h2o.doSafeREST(conn = conn, h2oRestApiVersion = h2oRestApiVersion,  : 
  Unexpected CURL error: Empty reply from server

Checking the h2o server logs, which you should do as well, will tell you:

10-08 20:11:57.165 192.168.0.4:54321     2125   #58072-18 INFO: Total file size: 1.81 GB
10-08 20:11:57.165 192.168.0.4:54321     2125   #58072-18 INFO: Parse chunk size 4194304
        onExCompletion for water.parser.ParseDataset$MultiFileParseTask@3588360e
        java.lang.OutOfMemoryError: Java heap space
:
:
Exception in thread "FJ-0-11" java.lang.OutOfMemoryError: Java heap space
2015-10-08 20:13:14.493:WARN:oejut.QueuedThreadPool:1 threads could not be stopped
10-08 20:13:23.033 192.168.0.4:54321     2125   FJ-0-5    ERRR: Out of Memory, Heap Space exceeded, increase Heap Size, from /192.168.0.4:54321
10-08 20:13:23.458 192.168.0.4:54321     2125   FJ-0-3    ERRR: Out of Memory, Heap Space exceeded, increase Heap Size, from /192.168.0.4:54321
10-08 20:13:23.033 192.168.0.4:54321     2125   FJ-0-13   ERRR: Out of Memory, Heap Space exceeded, increase Heap Size, from /192.168.0.4:54321
10-08 20:13:23.033 192.168.0.4:54321     2125   FJ-0-7    ERRR: Out of Memory, Heap Space exceeded, increase Heap Size, from /192.168.0.4:54321
10-08 20:13:26.541 192.168.0.4:54321     2125   FJ-0-5    FATAL: Exiting.
10-08 20:13:26.574 192.168.0.4:54321     2125   FJ-0-7    FATAL: Exiting.
10-08 20:13:26.575 192.168.0.4:54321     2125   FJ-0-3    FATAL: Exiting.
10-08 20:13:26.575 192.168.0.4:54321     2125   FJ-0-13   FATAL: Exiting.

I am running this on h2o Slater (3.2.0.5), so depending on your version, this may vary.