SCP plugin

Can you access other job's lastSuccessful?  I want to create a new scheduled job that grabs the last successful build artifact of a particular job.  Is this possible?

Does this plugin work on Windows with Cygwin installed? If so - this should probably be mentioned on this Confluence page.

If you only have a single artifact you want to copy somewhere, are there any thoughts to adding the ability to rename said artifact?  Near as I can tell, the destination only allows a relative directory path and there's no way to include a new file name.  This would be a very useful feature to me since I am using this tool to copy an artifact that has a version number in the file name (for QA and production) as well as for my development automated testing tool that would like to be version agnostic (so I don't have to change the automated tester every time the artifact version changes).

I'm thinking something like the following, but I haven't grabbed the source yet to find out what might be involved in doing this.

artifact name = product-1.0.ear

source: workspace/build/artifacts/*

destination: qa/product/1.0/product-1.0.ear

destination: dev/product/trunk/product-trunk.ear

(where product-*.ear is the file name, not a folder)

This plugin is very useful, it runs on a free-style software project but fails on a maven project.

Actually it never tries to run on a maven project, no log, no process.

Is there a solution for this?

The add files option doesn't seem to provide fields for me to enter data.  I have tried this on Firfox/IE and Chrome.  Is there example xml for the config.xml so I can sneak passed this?

I'm impressed, it's working like a charm! Still, I'd like to propose an enhancement:

I'd wish for an additional id field, so multiple destinations (i.e. repository paths and logins) can be set-up for the same host.

Right now, the hostname field serves two purposes:
1. Designate the FQDN (DSN-name or IP) of the target-host
2. State the name for a set of repository settings (i.e. port, repository path, user, password, etc.)

I have to admit that I don't use the SCP plugin to copy to a repository at all. Instead build artifacts are being copied to (Apache-) www-folders so they can then be downloaded and installed (.jads and .jars in my case). That's fine for a single target (i.e. repository), but appears to fail if there are numerous target directories (and each of them requiring a different login).

Should I a dare to open an enhancement issue?

I've got everything working brilliantly with this plugin except that it won't create subdirectories in the Source root Has anyone found a solution to this?

I'm trying to recursively copy over a directory from one location to another using the SCP plugin, and can't figure out how to do it with the Ant glob syntax.

Previously, I had a shell script that would do this:

scp -r $localFile $username@$remoteHostIP:$remoteDestDir

This would do a full recursive copy from "localFile" (really, local-dir) to the remoteDestDir location. I don't know how to specify this with the SCP plugin. I guess I could go back to running the shell script, or have several SCP calls that copy each sub-directory over, but that seems weird.

Any idea how I might do this?
-Dan.

Report an issue - SCP plug-in does not support concurrent SCP sessions.

As Hudson support multi-executor within one slave. it means scp stands chance of being invoked simultaneously by more than one job. I am encounted to such a requirement that success of job-0 triggers job-1 and job-2, both of which employ scp as post-build action. They failed scp simultaneously. but They do succeed separately.

I am willing to study your source codes. but I can not connect to https://svn.dev.java.net/svn/hudson/trunk/hudson/plugins/scp and I don't know why. can anyone give my a copy?

I am using this plugin in a maven project. It is working Excellent but Is there any option to run this plugin before running M2 extra build steps plugin.

Hi, I am a bit new to this plugin and I want to use the "key" authentication. How do I specify the key file or where do I put it? I tried using a url but it doesn't seem to work.

Hi, does this plugin has a log? where? I keep getting "Can't connect to server" for all external sites; hudson scp config page test can connect only to localhost; I also verified that I can connect to all these other sites with normal ssh client OK. Thank you.

[SOLVED] in this case I needed to provide following line  in sshd_config: PasswordAuthentication yes
despite this presumably being the default setting on this freebsd box and also working fine for manual logins. Cheers.

I have 2 question for the plugin developers.

1.This pluging appears in the last section of the project's configuration form and I need it to run it before a shell

script using shh pluging. Is it a way to change the order in which this pluging runs regards to others.

2. Right now is only possible to choose one site peer project.  Is it a way to include the option to add more sites

to upload to in the same project?.

"Can't connect to server"
Hudson is running as a war in glassfish, started by the user "glassfish".
I'm getting this error but can't determine why. When in cli on the box in which hudson is running, I can execute:
scp testFile root@remotebox:/tmp/
as the glassfish user and it works fine.

But when I try to add "remotebox" as an SCP host in hudson, I get "Can't connect to server".
I tried the fix above, editing sshd_config for PasswordAuthentication yes but it didn't seem to help.

Any other ideas?

Help me, help me, please help me, who can tell me what happened as below message...

COPY AND KEEP THE TREE?

currently, if you have on source:

dir1/dir2/dir3

and you specify:

source: "dir1/**"

destination "dir"

then all contents of all folders on the source "dir1/dir2/dir3" will be flattened (merge+overwrite) inside destination "dir";

how can you copy and preserve directory tree?

Nice  plugin,  it would be great if it allowed you to select two different repository   to  upload to.