Platform: Linux
TeamSite Version: 6.7.1

I was involved in a recent project where I was having webdeamon issues.  The service was not starting up, and in due course TeamSite interface could not be seen on the web.  The first thing I looked at was Apache.  However seem to be installed ok, and the service seem to be running ok.  After a thorough investigation of potential reasons why the service was failing to start we found the issue and the solution.

Issue:

The reason why the service fail to start was the nfslock service was disabled (not started in run level 345).

Solution:

Enable the nfslock service (rpc.statd and rpc.lockd)

Platform: Linux
TeamSite Version: 6.7.1
OpenDeploy Version: 6.2

I was involved in a project where I had just OpenDeploy 6.2 on a Linux environment.  However, the service would not start up. 
The error message appearing in od.log;

We had installed the prerequisite of the required library; libstdc++32-3.2.2-5.fdr.0.rh73.i386.rpm
However it seem to be clashing with other pre-installed libraries and applications on the server

Solution:

The Solution was to install the library in the OpenDeploy directory (e.g. /od-home/opendeployNG/lib), and thereafter modify
the OpenDeploy start-up service (iwod60 – note: make a backup copy of the service before making the changes below),
located in /etc/init.d/ with the following;

Install the library into the OpenDeploy location would not impact the server with clashes. 
Modfiying the start up service pointed to the library installed. After these changes were made,
I started the OpenDeploy service, and it worked!

Platform: Linux
TeamSite Version: 6.7.1
Content Store: Located on a NFS Share mounted on version 2

I was involved in a project where TeamSite reported the following warning in the iwtrace.logs

Warning: disk low    “/local/data/interwoven/iw-store/workflow”    Less than 50000 inodes (0 free of 0 inodes).

Cause:

This was diagnosed, and the following diagnosis was made

- The server reported a low number of inodes.
– ‘df -i’ showed that there were inodes at all (available, used and free) on the interwoven nfs share.
- This issue was triggered during diagnosis of a nfs lock (refer to content store lock article)
- netapp filer is confused when NFS share is mounted both as nfs v2 and nfs v3.

Solution:

Mounting the interwoven nfs share under nfs v3

Whilst running an OpenDeploy deployment that usually works no problem, I encountered the following below error.

"ERROR: IWDirent::Mkdir failed, err=30"
ERROR: /iwmnt/default/main/www/example/WORKAREA/Content 
could not be created. 
ERROR: Failed deploying local_directory[0] (/iwmnt/default/main
/test/WORKAREA/Content) to (/iwmnt/default/main/www/
example/WORKAREA/Content) 
ERROR: Failed 1 of 1 local_directories

SOLUTION

I created the target “Content” workarea in the above case and then this worked!

I think it is pretty obvious but someone / thing had deleted the “Content” workarea that OpenDeploy was configured to deploy too and so the “mkdir” was failing, presumably because of permissions.

If you try to create the directory that OpenDeploy is trying to deploy to this may resolve the above issue for you.

If you get the below issue when trying to edit or view a page in LiveSite then check whether the server has run out of disk space:

“An error occurred while getting the page”

In the servlet_out.log file you may also see the following:

"Caused by: com.interwoven.livesite.common.business.BusinessException: Failed to get page //snt2k020/default/main/UK/WORKAREA/Content/sites/one/home.page
 
        at com.interwoven.livesite.business.BusinessPage.getTemplate(BusinessPage.java:771) 
        at com.interwoven.livesite.business.BusinessPage.getPage(BusinessPage.java:636) 
        at com.interwoven.livesite.business.BusinessPage.verifyChecksum(BusinessPage.java:188) 
        ... 40 more 
Caused by: java.io.EOFException 
        at java.io.ObjectInputStream$PeekInputStream.readFully(ObjectInputStream.java:2228) 
        at java.io.ObjectInputStream$BlockDataInputStream.readShort(ObjectInputStream.java:2694)

and in the /var/adm/eventsubd.log you may see the following

"Error: CRITICAL: page header magic for block 1 not OK 0"

The most likely cause is a problem with the /iw-home/eventsubsystem/openjms.db file.

Move this file to another area (it may be very large and have taken up 100% disk space under iw-home – you can check this by doing a ‘df -k’ command and looking at the results)

Also delete openjms.lock and openjms.lg in the same folder.

Re-start TeamSite and it should re-create this file (it may be the same size when you “ls -la” but on disk iw-home is no longer at 100%.

We think this is due to the service “reserving” the space that was previously used.)

If you want to reduce the amount of “claimed” disk space you should delete any “.0″ files under “/iw-home/loacl/logs/iwevents” folder and restart TeamSite (/etc/init.d/iw.server stop) and then (/etc/init.d/iw.server start).

If you get the following error message when viewing a page in the LiveSite Runtime environment:

“Error 500: Filter [runtime.filter.OpenSessionsInView]: could not be initialized”

Please check the pagetype-config.xml file as this may have an error in it. This is usually located in the following folder:

…/WEB-INF/conf/livesite_customer

In case you have lost your working version of this file you can find one here.

If your experiencing trouble with OpenDeploy starting up and are seeing ‘ERROR: LicenseException: Bad host name’ errors in the OpenDeployNG/od.log file, this post will help resolve the issue.

To get your domain name and host name, on the Command Line type:

<Your OpenDeploy Home Directory>\OpenDeployNG\jre\bin>java -cp <Your OpenDeploy Home Directory>\opendeployng\lib\odng.jar com.interwoven.license.LicInfo -f <ODlic.log> -p OD

Use the output of this command to help generate the OpenDeploy license key

Now do the following:

On Windows check the following registry keys (Click Start, run, and type regedit)

\\HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\DhcpDomain
\\HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Domain

Make sure they match exactly, same string and case.  If dhcpDomain does not exit, create it (String value)

The key \\HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Domain holds the Primary DNS Domain, and is set as follows:

1. Right click My Computer and select properties
2. Select the ‘Computer Name’ tab
3. Click on the “More” button and populate the “Primary DNS Suffix”
4. A system reboot is required after the change.

Also the key \\HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\hostname holds the system hostname,  should be tied to the correct ip address, and defined in winnt\system32\drivers\etc\hosts.

To verify the hostname and IP “seen” by OD, you can review the odbase.log for the line “Network Address available on localhost” for the correct host or IP which should match the OD.lic file

On a Unix box, verify that the primary domain in /etc/resolv.conf matches the domain in /etc/hosts and the domain returned by nslookup on the server.

If your LiveSite page layout is missing you will see the following error on the runtime website:

Error 500: Filter [runtime.filter.DispatchFilter]: filter is unavailable.

To fix this copy your page layout file to the following folder:

…WEB-INF/classes/com/interwoven/xsl/runtime/

This page is created to hold Common OpenDeploy errors with their possible causes.

It was originally based on a Interwoven Support article but I have updated and extended this to include more and newer errors.

Where there is more than one possible cause, this is noted as #1 and #2 and so forth.

ERROR: outer-loop send failure, err=32 [Unknown socket error]
ERROR: recv failure, err=131 [Unknown socket error]
 Possible Cause#1: Firewall - make sure port is open
 Possible Cause#2: Need to put timeout=25000 (or more)
at top of both server and client config files

ERROR: Failed applying access to (path_to_file)
ERROR: Could not get file data(./filename) from socket.
 Possible Cause:Need to run OD as root

Error: Can Not read Version String Response
 Possible Cause: Port already in use or not open
Possible Cause: If you have multiple deployments to the same
server using the same port number, you will get this message.
In order to deploy simultaneously to the same server,
you need to run each deployment on different ports.

ERROR: Could not read src-dir ack from client.
 Possible Cause: Seems to have many causes including permissions

ERROR: Could not read directive.
 Possible Cause: problem with a script called by OpenDeploy

Error: Ambiguous cross-platform name comparison
 Possible Cause: On solaris to NT if there are files with same name
but different case. Need to change the name.

Error: 'c:\Program' is not recognized as an internal
or external command, operable program or batch file.
 Possible Cause: Need to set $IWODHOME to
Progra1~ from Program Files

ERROR: Could not write content size to client
ERROR: Could not write status to client
 Possible Cause: Using different versions of OD

ERROR: Could not read check-keyfile result.
 Possible Cause: Using wrong version of OD

Error: Outer loop send failure. Err 10053.
Cannot return status to server. Could not authenticate client
 Possible Cause: Usually because of the names put into the .cfg files.
Double check the names or use the ip numbers

ERROR: User is not authorized for this deploy!
ERROR: Could not authenticate client.
 Possible Cause: OpenDeploy not running as root on Solaris or
Local System on Windows

ERROR: Could not read deployment name.
 Possible Cause: Check to see if the the OD versions are the same on
both server and client (iwdeploy -v)

ERROR: Could not find matching OpenDeploy client for (ServerX)
 Possible Cause: Configuration has a host name instead of an IP address
set up for deployment

ERROR: Could not rename
 Possible Cause: Disk space on target server

ERROR: Could not read content size from remote.
 Possible Cause: Firewall timeout

ERROR: Could not read server status, client OK
 Possible Cause: Firewall timeout

ERROR: Could not read commit complete
 Possible Cause: Firewall timeout

ERROR: Failed 1 of 1 local_directories
 Possible Cause: Firewall timeout

ERROR: Could not read rollback start state
 Possible Cause: Firewall timeout

ERROR: Unable to do hand shake with the sender
 Possible Cause: Firewall timeout

ERROR: Could not write content size to client.
 Possible Cause: Firewall timeout

ERROR: Could not write status to client
 Possible Cause: Firewall timeout

ERROR: outer-loop send failure, fd=4, err=32 [Unknown socket error]
 Possible Cause: Firewall timeout

ERROR: not an allowed deployment path
 Possible Cause: NAT source IP must be configure in odrcvr.xml

ERROR: Failed to locate Access Service,
 Possible Cause: Update the OD-conf/framework.properties file
to reflect correct server name