nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

NEED HELP: Publish OD failed

Constantin

Hi All,

Need help urgently, currently we are using Teamsite, OD, and Livesite 7.3

For all the server we are using Windows Server 2008 64 bit R2.
We have 1 Teamsite server and 2 Runtime Server, and configured the deployment to be fan-out deploy, so it deploy to both runtime server when publishing to PLC workflow.

Now we get this error when publishing in the livesite log:

LIB: 2012-06-01 10:18:55 ERROR: MoveFile Failed while moving from (C:\Interwoven\OpenDeployNG\tmp\temp_files\interwoven\livesitedisplayservices\runtime\web\iwov-resources\img\2-.jpg.iwnew) to (C:\interwoven\livesitedisplayservices\runtime\web\iwov-resources\img\2-.jpg), errno=5
LIB: 2012-06-01 10:18:55 ERROR: Failed rename of (C:\Interwoven\OpenDeployNG\tmp\temp_files\interwoven\livesitedisplayservices\runtime\web\iwov-resources\img\2-.jpg.iwnew) to (C:\interwoven\livesitedisplayservices\runtime\web\iwov-resources\img\2-.jpg) -- GetLastError()=998

this error happened when we are trying to publis a page with a lot of dependency files (> 80 files).

If we try to publish a file that only a page or publish only an image, the publish process is completed. It publish correctly to both server.

I've attached the log file and config file that i use in our production and development server.

Thanks before,
Constantin

Find more posts tagged with

Comments

Constantin

Just an update,

If we try the following scenario:
- Publish a page with a resource that throw an error (in this case 2-.jpg) --> Failed
- Publish the resource it self 2-.jpg --> Successful
- Publish a page without the resource --> Successful
- Publish the same resource but rename the file --> successful for one time, then for the next publish it will return error...

Is someone have encountered the same error ?

Regards,
Constantin

nipper

Sometimes, if a transactional deployment fails badly, it leaves files named .iwnew or some such on the target system.

Can you log on and view the files on the target server ?

Also, is OD running as root/admin on the target ?

Rick Poulin

Also, is OD running as root/admin on the target ?

More generally, make sure that the user running OD on the target has sufficient permissions to delete 2-.jpg, and that there isn't a process holding a lock on that file and preventing it from being removed.

Much less likely for this specific rename, but it's also remotely possible that your disk is full (typically is more to blame for the copy of 2-.jpg to 2-.jpg.iwold, or the creation of the 2-.jpg.iwnew in the first place..)

Constantin

Hi All,

Thanks for your reply,

@rpoulin:
Yes we already make sure that the permission to access / delete the file is granted.

As you see in my earlier posting some of the page is successful published.

Some times if you try to rename the file, it become successful, but then the second or third time publish for the same page, it will return an error again.

@nipper:
We have 200 GB free space, and the web it self doesn't used the space that much, i think it's less than 1% (around 400 MB or less).

I've try to set the deployment to retry the error 10 times but still doesnt help much.

I've log the case to the support but they also still finding the error, and they ask us to get the professional services to consult. They said standard PLC workflow doesn't support fan out deployment.

Screen Shot 2016-10-14 at 21.10.47.png

Rick Poulin

They said standard PLC workflow doesn't support fan out deployment.

O rly? You might want to tell support to RTFM. SiteDeveloper guide appendix E.

nipper

O rly? You might want to tell support to RTFM. SiteDeveloper guide appendix E.

That is not a fanout deployment, that adds additional environment but not fanouts.

Still changing that is not horribly difficult. Recent I changed the OD config from a server deployment to a deployment group.

DisplayTopN_onChart.png

Constantin

Hi All,

Finally we have found on what might cause the problem.

our server currently using Apache as proxy server and load balancer (future use), so Apache as the front gate and then livesite at the back.

So when we tried to turn of the apache and perform the publish process, it turns out that it's success. Now we are trying to find out why Apache is locking certain files, anybody know what and how to solve this problem ?

Regards,
Constantin

Topic_39668__OS BIRT Chart Label on Last Data Point.rptdesign

OS BIRT Chart with last point label.png

Rick Poulin

How do you have it configured w.r.t. proxy? If it's a full proxy (i.e. proxy / to :1776/) then there's no reason for it to lock any files deployed into LSDS. Are you perhaps only proxying .page requests, and serving up static files as a virtual directory? Are you using a RewriteMap+RewriteLock?

Constantin

How do you have it configured w.r.t. proxy? If it's a full proxy (i.e. proxy / to :1776/) then there's no reason for it to lock any files deployed into LSDS. Are you perhaps only proxying .page requests, and serving up static files as a virtual directory? Are you using a RewriteMap+RewriteLock?

We use proxy for user request to internal application. We use ajp from apache to LSDS.

Are the apache cache config may lock the file?

Rick Poulin

I've never heard of AJP locking files in the web application root, but it may prevent you from hot-redeploying apps...

Which files are getting locked and when? How have you determined that Apache is the culprit?

Constantin

Hi All,

Just an update, finally we know and able to solve the problem.
This configuration in alias.conf that makes the publish process return errno=5

Alias /iwov-resources "C:/Interwoven/LiveSiteDisplayServices/runtime/web/iwov-resources"

Options -Indexes FollowSymLinks MultiViews ExecCGI
AllowOverride None
Order allow,deny
Allow from all

So for now we comment this configuration, and the publishing process now is ok.

But there is another issue coming other than this issue, sometimes the PLC flow is stuck on Activates optimost experiments task. There is no error found in the logs. Although the files is published and commit deploy process is completed, but the workflow itself never completed.