Process 'command 'bash'' finished with non-zero exit value 1
Yang Luo
ke...@0xdata.com
with respect to the message
"Error occurred during initialization of VM
Could not reserve enough space for object heap"
we typically code on laptops with 16GB of dram, and the gradlew build and test is optimized to be aware of that as an upper limit.
I'd have to check what the minimum dram requirement is when running the test step in gradlew.
I'm assuming you are using a laptop? Do you know how much dram it has? I suspect maybe only 4GB?
You can skip the test phase of the gradlew build by using
gradlew build -x test
Also: I see you are using windows 7. Are you running the build with cygwin or native windows?
-kevin
"
Yang Luo
On Monday, June 15, 2015 at 6:39:34 PM UTC+8, ke...@0xdata.com wrote:
Hi kevin,
If I want to run gradlew build command, my laptop should have least 16G dram, is that right？
Now I change the enviroment to unbutu14.04.
-kevin
"
ke...@0xdata.com
> Now I change the enviroment to unbutu14.04.
> and I do run gradlew build -x test
> there a error: if you have time, please give me some advice. Thanks.
>
>
>
>
>
> /home/guow/GitHUb/h2o-3/h2o-py/h2o/__init__.py:h2o-py:setProjectVersion took 0.017 secs
> :h2o-py:buildDist
> usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
> or: setup.py --help [cmd1 cmd2 ...]
> or: setup.py --help-commands
> or: setup.py cmd --help
>
>
> error: invalid command 'bdist_wheel'
> :h2o-py:buildDist FAILED
> :h2o-py:buildDist took 0.217 secs
>
Because of various issues associated with the python transition to using wheel instead of distutils, for distributing packages, I have seen a lot of odd things in people's setups.
I have a general recipe I recommend, here:
https://groups.google.com/forum/#!searchin/h2ostream/So$20install$20issues$20on$20fresh|sort:relevance/h2ostream/eWX5m_QtNvA/b06qTl5ZJykJ
the reply after that, I mention to reinstall wheel.
Try reading what I say there, and try it (I use ubuntu exclusively..12.04.5 with trusty backport. I have two ubuntu 14.04 systems, but haven't done the h2o package on them).
You may just have not installed wheel, but I've seen many problems with wheel not installed correctly and difficult to debug. So I would strong recommend doing the step by step I have in that link above, and report your results here.
thanks!
-kevin
ke...@0xdata.com
I said I'd also check how much dram the gradlew build uses when enabled to run the tests.
We do tests that are both single h2o process, and multiple h2o processes, to emulate h2o in a cloud.
I looked at one that I think is the biggest consumer of memory, the h2o-algos section of the junit tests that are run.
It starts 5 h2o jvms, each with 3 gbyte heap. So that's 15GB
that means you need a 16GB dram machine (laptops can have 16GB nowadays)
the h2o stdout that's produced during this test phase shows the cloud of 5, each with 3G heaps. 5*3 = 15, and you want a little more than that.
I think we should have this requirement listed on the README.
06-16 00:16:54.034 172.17.2.152:44008 1202 main INFO: Java heap totalMemory: 2.88 GB
06-16 00:16:54.035 172.17.2.152:44008 1202 main INFO: Java heap maxMemory: 2.88 GB
06-16 00:16:54.035 172.17.2.152:44008 1202 main INFO: Java version: Java 1.7.0_80 (from Oracle Corporation)
06-16 00:16:54.035 172.17.2.152:44008 1202 main INFO: JVM launch parameters: [-ea, -Xmx3g, -Xms3g, -Dai.h2o.name=junit_cluster_1168, -Dai.h2o.baseport=44000, -Dai.h2o.ga_opt_out=yes]
06-16 00:16:54.035 172.17.2.152:44008 1202 main INFO: OS version: Linux 3.13.0-54-generic (amd64)
06-16 00:16:54.035 172.17.2.152:44008 1202 main INFO: Machine physical memory: 125.82 GB
06-16 00:16:54.036 172.17.2.152:44008 1202 main INFO: X-h2o-cluster-id: 1434439013776
06-16 00:16:54.036 172.17.2.152:44008 1202 main INFO: Opted out of sending usage metrics.
06-16 00:16:54.036 172.17.2.152:44008 1202 main INFO: Possible IP Address: vethOEDGYM (vethOEDGYM), fe80:0:0:0:fc40:96ff:fe81:dcd8%6
06-16 00:16:54.036 172.17.2.152:44008 1202 main INFO: Possible IP Address: eth0 (eth0), fe80:0:0:0:ec4:7aff:fe34:5032%2
06-16 00:16:54.036 172.17.2.152:44008 1202 main INFO: Possible IP Address: br0 (br0), fe80:0:0:0:ec4:7aff:fe34:5032%4
06-16 00:16:54.037 172.17.2.152:44008 1202 main INFO: Possible IP Address: br0 (br0), 172.17.2.152
06-16 00:16:54.037 172.17.2.152:44008 1202 main INFO: Possible IP Address: lo (lo), 0:0:0:0:0:0:0:1%1
06-16 00:16:54.037 172.17.2.152:44008 1202 main INFO: Possible IP Address: lo (lo), 127.0.0.1
06-16 00:16:54.037 172.17.2.152:44008 1202 main INFO: Internal communication uses port: 44009
06-16 00:16:54.037 172.17.2.152:44008 1202 main INFO: Listening for HTTP and REST traffic on http://172.17.2.152:44008/
06-16 00:16:54.038 172.17.2.152:44008 1202 main INFO: H2O cloud name: 'junit_cluster_1168' on /172.17.2.152:44008, discovery address /235.9.159.184:60169
06-16 00:16:54.038 172.17.2.152:44008 1202 main INFO: 2. Point your browser to http://localhost:55555
06-16 00:16:54.120 172.17.2.152:44008 1202 main INFO: Log dir: '/tmp/h2o-jenkins/h2ologs'
06-16 00:16:54.120 172.17.2.152:44008 1202 main INFO: Cur dir: '/home4/jenkins/slave_dir_from_mr-0xa1/workspace/master_gradle_build/h2o-algos'
06-16 00:16:54.125 172.17.2.152:44008 1202 main INFO: HDFS subsystem not available
06-16 00:16:54.127 172.17.2.152:44008 1202 main INFO: S3 subsystem successfully initialized
06-16 00:16:54.127 172.17.2.152:44008 1202 main INFO: Flow dir: '/home4/jenkins/h2oflows'
06-16 00:16:54.145 172.17.2.152:44008 1202 main INFO: Cloud of size 1 formed [/172.17.2.152:44008]
06-16 00:16:57.136 172.17.2.152:44008 1202 FJ-126-15 INFO: Cloud of size 5 formed [/172.17.2.152:44000, /172.17.2.152:44002, /172.17.2.152:44004, /172.17.2.152:44006, /172.17.2.152:44008]