[
https://issues.apache.org/jira/browse/VFS-546?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14220758#comment-14220758
]
Sergio Fernández commented on VFS-546:
--------------------------------------
I'll try to debug with some of those tools...
I do not understand, how [the same
test|https://bitbucket.org/mico-project/platform/src/db336d8b7347d03a4f980f8c89e8e17fb57cc33a/api/java/persistence/src/test/java/eu/mico/platform/persistence/test/MarmottaContentTest.java?at=master#cl-45]
has such different figures remotely than local:
{quote}
{code:none}
sergio@wolfcastle:~/projects/mico/platform/api/java/persistence$ mvn test
-Dtest=MarmottaContentTest#testFramework105 -Dtest.host=192.168.56.101
...
Running eu.mico.platform.persistence.test.MarmottaContentTest
INFO [main] (MarmottaContentTest.java:52) - Using
ftp://mico:[email protected]/2799e422-8d99-4330-9765-4fb388e96594/7403e783-30d0-4338-8d19-769b489e6b3c.bin
as path for testing
INFO [main] (VfsLog.java:138) - Using "/tmp/vfs_cache" as temporary files
store.
DEBUG [main] (MarmottaContentTest.java:58) - initializing VFS Manager: 86ms
DEBUG [main] (MarmottaContentTest.java:62) - resolving content item path: 69ms
DEBUG [main] (MarmottaContentTest.java:66) - checking parent directory: 51ms
DEBUG [main] (MarmottaContentTest.java:70) - creating parent directory: 24ms
DEBUG [main] (MarmottaContentTest.java:75) - resolving content part path: 0ms
DEBUG [main] (MarmottaContentTest.java:79) - creating content part file: 25ms
DEBUG [main] (MarmottaContentTest.java:83) - getting output stream: 3ms
DEBUG [main] (MarmottaContentTest.java:88) - writing content to the stream:
2265ms
Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 2.936 sec
{code}
{quote}
{quote}
{code:none}
mico@mico:~/src/platform/api/java/persistence$ mvn clean test
-Dtest=MarmottaContentTest#testFramework105 -Dtest.host=localhost
...
Running eu.mico.platform.persistence.test.MarmottaContentTest
INFO [main] (MarmottaContentTest.java:52) - Using
ftp://mico:mico@localhost/a18ba526-0a34-4d73-9b7d-2144eabd9ace/ee2a6b28-43f0-4535-b663-f38c95863a0c.bin
as path for testing
INFO [main] (VfsLog.java:138) - Using "/tmp/vfs_cache" as temporary files
store.
DEBUG [main] (MarmottaContentTest.java:58) - initializing VFS Manager: 159ms
DEBUG [main] (MarmottaContentTest.java:62) - resolving content item path: 76ms
DEBUG [main] (MarmottaContentTest.java:66) - checking parent directory: 60ms
DEBUG [main] (MarmottaContentTest.java:70) - creating parent directory: 25ms
DEBUG [main] (MarmottaContentTest.java:75) - resolving content part path: 0ms
DEBUG [main] (MarmottaContentTest.java:79) - creating content part file: 40ms
DEBUG [main] (MarmottaContentTest.java:83) - getting output stream: 3ms
DEBUG [main] (MarmottaContentTest.java:88) - writing content to the stream:
88767ms
Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 90.01 sec
{code}
{quote}
Maybe something is wrng in [our
configuration|https://bitbucket.org/mico-project/platform/src/db336d8b7347d03a4f980f8c89e8e17fb57cc33a/api/java/persistence/src/main/java/eu/mico/platform/persistence/util/VFSUtils.java?at=master#cl-18],
[~b.eckenfels]? And for sure we can discard a server-sice issue, because
alternative implementations (curl and c++) work as expected.
> VFS very slow resolving FTP files
> ---------------------------------
>
> Key: VFS-546
> URL: https://issues.apache.org/jira/browse/VFS-546
> Project: Commons VFS
> Issue Type: Bug
> Affects Versions: 2.0
> Environment: Debian
> Reporter: Sergio Fernández
> Priority: Minor
>
> We have found a performance issue using VFS to access files via FTP. Pushing
> few MBs takes too long... The code is:
> {code:java}
> FileSystemManager fsmgr = VFS.getManager();
> FileObject f = fsmgr.resolveFile(path);
> if(d.exists()) {
> return f.getContent().getOutputStream();
> }
> {code}
> With the same FTP server both using {{curl}} and a C++ implementation the
> same file requires few seconds.
> Could be a configuration issue, a usage issue or a bug?
> Thanks in advance.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
