Cannot wipe workspace, blocked by SCM

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Cannot wipe workspace, blocked by SCM

tfroyland

I am trying to “Wipe Out Workspace” and am receiving the message:

 

Error: Wipe Out Workspace blocked by SCM

The SCM for this project has blocked this attempt to wipe out the project's workspace.

 

I am using Perforce as my SCM, and the polling and building seems to be working fine. In the Hudson.log I see the following:

 

Feb 26, 2010 2:03:49 PM hudson.plugins.perforce.PerforceSCM processWorkspaceBeforeDeletion

INFO: Veto workspace cleanup

 

Hopefully, someone can answer one of the following two questions. Please give a try to either…

 

1) Why is this happening / how can I fix it? I’ve seen a post in the archives with this exact problem with no solution, so I am not hopeful on this point.

 

2) How can I get more information on what Perforce command is being executed and what error is being returned? I have added a system log with the logger “Hudson.plugins.perforce”, and I am seeing the same info that is in the log, even though I specified “all” in the logger configuration.

 

Some additional information: I am also seeing some Perforce errors in the log that indicate that the Perforce process terminated suddenly during a polling, although the polling seems to be working fine. These errors don’t seem to be related to the wipe of the workspace based on the timings I’ve been able to do.

 

Any suggestions for tracking down this problem would be dearly appreciated.

 

Thanks.

 

Reply | Threaded
Open this post in threaded view
|

Re: Cannot wipe workspace, blocked by SCM

carlq
The Hudson plugin for Perforce is blocking this request. Part of the Hudson API for SCM plugins is this interaction to allow the SCM plugin to veto a wipe. There is no p4 command issue or real failure happening here.

Because Perforce keeps server state, the plugin can't simply allow a local workspace wipe to happen unless it can inform the server somehow. That code hasn't yet been written, and even when it is, I suspect it will be limited to cases where hudson is managing the workspace.

This has been on my mind too, so I've created a Jira issue to track this improvement: http://issues.hudson-ci.org/browse/HUDSON-5777


On Fri, Feb 26, 2010 at 11:11 AM, Froyland, Todd <[hidden email]> wrote:

I am trying to “Wipe Out Workspace” and am receiving the message:

 

Error: Wipe Out Workspace blocked by SCM

The SCM for this project has blocked this attempt to wipe out the project's workspace.

 

I am using Perforce as my SCM, and the polling and building seems to be working fine. In the Hudson.log I see the following:

 

Feb 26, 2010 2:03:49 PM hudson.plugins.perforce.PerforceSCM processWorkspaceBeforeDeletion

INFO: Veto workspace cleanup

 

Hopefully, someone can answer one of the following two questions. Please give a try to either…

 

1) Why is this happening / how can I fix it? I’ve seen a post in the archives with this exact problem with no solution, so I am not hopeful on this point.

 

2) How can I get more information on what Perforce command is being executed and what error is being returned? I have added a system log with the logger “Hudson.plugins.perforce”, and I am seeing the same info that is in the log, even though I specified “all” in the logger configuration.

 

Some additional information: I am also seeing some Perforce errors in the log that indicate that the Perforce process terminated suddenly during a polling, although the polling seems to be working fine. These errors don’t seem to be related to the wipe of the workspace based on the timings I’ve been able to do.

 

Any suggestions for tracking down this problem would be dearly appreciated.

 

Thanks.

 


Reply | Threaded
Open this post in threaded view
|

RE: Cannot wipe workspace, blocked by SCM

Curt Patrick

An alternative that might work for you:  Instead of wiping out the workspace, do your build activities in some directory parallel to the workspace.  You can wipe that out whenever you like (I presume that the Perforce plugin will have no objection), then copy the workspace over there.  Since you are not building in the workspace itself it should never need wiped out?

 

Curt

 

From: Carl Quinn [mailto:[hidden email]]
Sent: Friday, February 26, 2010 12:03 PM
To: [hidden email]
Subject: Re: Cannot wipe workspace, blocked by SCM

 

The Hudson plugin for Perforce is blocking this request. Part of the Hudson API for SCM plugins is this interaction to allow the SCM plugin to veto a wipe. There is no p4 command issue or real failure happening here.

 

Because Perforce keeps server state, the plugin can't simply allow a local workspace wipe to happen unless it can inform the server somehow. That code hasn't yet been written, and even when it is, I suspect it will be limited to cases where hudson is managing the workspace.

 

This has been on my mind too, so I've created a Jira issue to track this improvement: http://issues.hudson-ci.org/browse/HUDSON-5777

 

On Fri, Feb 26, 2010 at 11:11 AM, Froyland, Todd <[hidden email]> wrote:

I am trying to “Wipe Out Workspace” and am receiving the message:

 

Error: Wipe Out Workspace blocked by SCM

The SCM for this project has blocked this attempt to wipe out the project's workspace.

 

I am using Perforce as my SCM, and the polling and building seems to be working fine. In the Hudson.log I see the following:

 

Feb 26, 2010 2:03:49 PM hudson.plugins.perforce.PerforceSCM processWorkspaceBeforeDeletion

INFO: Veto workspace cleanup

 

Hopefully, someone can answer one of the following two questions. Please give a try to either…

 

1) Why is this happening / how can I fix it? I’ve seen a post in the archives with this exact problem with no solution, so I am not hopeful on this point.

 

2) How can I get more information on what Perforce command is being executed and what error is being returned? I have added a system log with the logger “Hudson.plugins.perforce”, and I am seeing the same info that is in the log, even though I specified “all” in the logger configuration.

 

Some additional information: I am also seeing some Perforce errors in the log that indicate that the Perforce process terminated suddenly during a polling, although the polling seems to be working fine. These errors don’t seem to be related to the wipe of the workspace based on the timings I’ve been able to do.

 

Any suggestions for tracking down this problem would be dearly appreciated.

 

Thanks.

 

 

Reply | Threaded
Open this post in threaded view
|

RE: Cannot wipe workspace, blocked by SCM

tfroyland

Carl,

 

Thanks for the explanation. Now I know I can stop trying to “fix” this and start figuring out how to accomplish it another way, such as manually wiping the directory and setting “force refresh of workspace” in the configuration. Too bad the Perforce plug-in can’t just set this “force” flag when someone requests a wipe.

 

Curt,

 

An interesting idea. We use Maven for the build, and while I’m reasonably sure we could do this, I have no idea how. I don’t want to mess up the pom files such that our developers won’t be able to use them for their own builds. Perhaps when I have the time I will look into this a bit.

 

Appreciate all your answers.

 

-Todd

 

From: Curt Patrick [mailto:[hidden email]]
Sent: Friday, February 26, 2010 12:08 PM
To: [hidden email]
Subject: RE: Cannot wipe workspace, blocked by SCM

 

An alternative that might work for you:  Instead of wiping out the workspace, do your build activities in some directory parallel to the workspace.  You can wipe that out whenever you like (I presume that the Perforce plugin will have no objection), then copy the workspace over there.  Since you are not building in the workspace itself it should never need wiped out?

 

Curt

 

From: Carl Quinn [mailto:[hidden email]]
Sent: Friday, February 26, 2010 12:03 PM
To: [hidden email]
Subject: Re: Cannot wipe workspace, blocked by SCM

 

The Hudson plugin for Perforce is blocking this request. Part of the Hudson API for SCM plugins is this interaction to allow the SCM plugin to veto a wipe. There is no p4 command issue or real failure happening here.

 

Because Perforce keeps server state, the plugin can't simply allow a local workspace wipe to happen unless it can inform the server somehow. That code hasn't yet been written, and even when it is, I suspect it will be limited to cases where hudson is managing the workspace.

 

This has been on my mind too, so I've created a Jira issue to track this improvement: http://issues.hudson-ci.org/browse/HUDSON-5777

 

On Fri, Feb 26, 2010 at 11:11 AM, Froyland, Todd <[hidden email]> wrote:

I am trying to “Wipe Out Workspace” and am receiving the message:

 

Error: Wipe Out Workspace blocked by SCM

The SCM for this project has blocked this attempt to wipe out the project's workspace.

 

I am using Perforce as my SCM, and the polling and building seems to be working fine. In the Hudson.log I see the following:

 

Feb 26, 2010 2:03:49 PM hudson.plugins.perforce.PerforceSCM processWorkspaceBeforeDeletion

INFO: Veto workspace cleanup

 

Hopefully, someone can answer one of the following two questions. Please give a try to either…

 

1) Why is this happening / how can I fix it? I’ve seen a post in the archives with this exact problem with no solution, so I am not hopeful on this point.

 

2) How can I get more information on what Perforce command is being executed and what error is being returned? I have added a system log with the logger “Hudson.plugins.perforce”, and I am seeing the same info that is in the log, even though I specified “all” in the logger configuration.

 

Some additional information: I am also seeing some Perforce errors in the log that indicate that the Perforce process terminated suddenly during a polling, although the polling seems to be working fine. These errors don’t seem to be related to the wipe of the workspace based on the timings I’ve been able to do.

 

Any suggestions for tracking down this problem would be dearly appreciated.

 

Thanks.

 

 

Reply | Threaded
Open this post in threaded view
|

Clean Builds/Perforce SCM

Steve M. Robbins
In reply to this post by Curt Patrick
Hi,

I have a similar need to the original poster that is also
apparently frustrated by the interaction of Perforce and Hudson.

Our need is to do a nightly "clean build"; i.e. wipe out the workspace,
re-sync, and build.  Since the Perforce SCM plugin doesn't have
the option to do this (c.f. HUDSON-5182 and HUDSON-3966), we've resorted
to a simple script that:

1. removes all files in the workspace
2. runs "p4 sync -f"

Unfortunately, we have to run this script as the first step in the
build process -- after the Hudson-generated sync -- so we pull changed
files twice, etc.  Not a very efficient solution.  Has anyone a better
idea?


On Fri, Feb 26, 2010 at 12:08:12PM -0800, Curt Patrick wrote:
> An alternative that might work for you:  Instead of wiping out the
> workspace, do your build activities in some directory parallel to the
> workspace.  You can wipe that out whenever you like (I presume that the
> Perforce plugin will have no objection), then copy the workspace over
> there.  Since you are not building in the workspace itself it should
> never need wiped out?

That's a great idea.  If I had a sane build system, I'd do exactly
that.  Sadly, I'm stuck with Microsoft Visual Studio which not only
wants to build in subdirs of the sources but also generates source
files (from resource files) in the source tree.

Thanks,
-Steve

signature.asc (197 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Out of Office AutoReply: Clean Builds/Perforce SCM

pcampbell
This post has NOT been accepted by the mailing list yet.
Out of Office AutoReply: Clean Builds/Perforce SCM

I am out of the office Monday, March 1 through Friday, March 5 and am unable to respond to your email.  I will return on Monday, March 8.

If you need immediate assistance for Hudson issues, contact Jason Collins, otherwise please contact Juan Nunez.