CM Repository problem

Hello all,
I have a CM repository set up with FSDB persistence mode. The files sit on a Windows server other than the portal server. The server that these files sit on had been shutdown over the weekend for some reason. When I came in this morning, the CM repository, of course, could not see the files. We have since brought up the server where the files sit, but KM still cannot see the files. In the KM monitoring log, I see the error:
Startup Error: folder '
10.0.32.191\PRDRepositoryEP6\san\Files' does not exist
What can I do to get the portal to see this again?
Thanks!
-Stephen Spalding
Web Developer
Graybar

I solved the problem on my own...I was able to get the repository to come back without restarting the portals.
What I did was create a new network path that had all of the same attributes as the network path used for the repository in question. I didn't have to tell the repository to use the new network path, the repository just started working after I created it. It's almost like the portal had to be reminded of where the repository's files were located.
The reason why I remembered to try this is because I had encountered the same problem in our dev env last week and in the process of trying to create a new repository, I noticed that the old one started working after I created a new network path.
Thanks to those who responded,
respectfully,
Stephen Spalding

Similar Messages

OVM 2.2 and NFS Repository Problems

Hi All.
I have recently started trying to upgrade our installation to 2.2
but have run into a few problems, mostly relating to the different
way that storage repositories are handled in comparison to 2.1.5.
We use NFS here to provide shared storage to the pools.
I wanted to setup a new two node server pool (with HA), so I upgraded
one of the servers to from 2.1.5 to 2.2 to act as pool master. That
worked ok and this server seems to be working fine in isolation:
master# /opt/ovs-agent-2.3/utils/repos.py -l
[ * ] 865a2e52-db29-48f1-98a0-98f985b3065c => augustus:/vol/OVS_pv_vpn
master# df /OVS
Filesystem 1K-blocks Used Available Use% Mounted on
augustus:/vol/OVS_pv_vpn
47185920 16083008 31102912 35% /var/ovs/mount/865A2E52DB2948F198A098F985B3065C
(I then successfully launched a VM on it.)
The problem is when I try to add a second server to the pool. I did
a fresh install of 2.2 and configured the storage repository to be the
same as that used on the first node:
vm1# /opt/ovs-agent-2.3/utils/repos.py --new augustus:/vol/OVS_pv_vpn
vm1# /opt/ovs-agent-2.3/utils/repos.py -r 865a2e52-db29-48f1-98a0-98f985b3065c
vm1# /opt/ovs-agent-2.3/utils/repos.py -l
[ R ] 865a2e52-db29-48f1-98a0-98f985b3065c => augustus:/vol/OVS_pv_vpn
When I try to add this server into the pool using the management GUI, I get
this error:
OVM-1011 Oracle VM Server 172.22.36.24 operation HA Check Prerequisite failed: failed:<Exception: ha_precheck_storage_mount failed:<Exception: /OVS must be mounted.> .
Running "repos.py -i" yields:
Cluster not available.
Seems like a chicken and egg problem: I can't add the server to the pool without a
mounted /OVS, but mounting /OVS is done by adding it to the pool? Or do I have that
wrong?
More generally, I'm a bit confused at how the repositories are
supposed to be managed under 2.2.
For exaple, the /etc/init.d/ovsrepositories script is still present,
but is it still used? When I run it, it prints a couple of errors and
doesn't seem to mount anything:
vm1# service ovsrepositories start
/etc/ovs/repositories does not exist
Starting OVS Storage Repository Mounter...
/etc/init.d/ovsrepositories: line 111: /etc/ovs/repositories: No such file or directory
/etc/init.d/ovsrepositories: line 111: /etc/ovs/repositories: No such file or directory
OVS Storage Repository Mounter Startup: [ OK ]
Should this service be turned off? It seem that ovs-agent now takes
responsibility for mounting the repositories.
As an aside, my Manager is still running 2.1.5 - is that part of the
problem here? Is it safe to upgrade the manager to 2.2 while I still
have a couple of pools running 2.1.5 servers?
Thanks in adavance,
Robert.

rns wrote:
Seems like a chicken and egg problem: I can't add the server to the pool without a
mounted /OVS, but mounting /OVS is done by adding it to the pool? Or do I have that
wrong?You have that wrong -- the /OVS mount point is created by ovs-agent while the server is added to the pool. You just need access to the shared storage.
For exaple, the /etc/init.d/ovsrepositories script is still present,
but is it still used?No, it is not. ovs-agent now handles the storage repositories.
As an aside, my Manager is still running 2.1.5 - is that part of the
problem here? Yes. You absolutely need to upgrade your Manager first to 2.2 before attempting to create/manage a 2.2-based pool. The 2.1.5 Manager doesn't know how to tell the ovs-agent how to create/join a pool properly. The upgrade process is detailed in [the ULN FAQ|https://linux.oracle.com/uln_faq.html#10].

OVM 3.0.1 local repository problem

Good morning all, i am really new in OVM and i am facing a big issue that stops me evaluating this product.
I have a couple of servers, connected to a S.A.N. array. I can see from both the servers i added to a clustered pool, and i am able to create a shared repository without problems.
I am not able to see local disks in the OVM manager administration and therefore i can't create local repositories. I tried all i found in this forum, but without success.
Let's focus on server1: it has a couple of 146GB disks. I used one of them for OVS installation leaving the second disk alone, without partitioning it.
Tried to create local repository in the clustered pool, but no way...
So i created a single full-disk partition and retried to create repo: still no way
Then i created an ocfs2 filesystem in the new partition but, again, i couldnt see physical local server1 disk.
Every time i changed partitions configuration, i obviously did rescanning of physical disks.
I all my tests, local physical disks selection list in Generic Local Storage Array @ node1 is always empty.
Any hint about solving this issue? Any good pointer to an hands-on guide (official docs are not so good)? Any suggestion about what to look at in log files for debugging?
Any answer is welcome...
Thank you all!

I was able to do this as follows
1. have an untouched unformatted disk (no partitions, no file system)
2. in hardware under the vmserver name , scan for the disk and it should show in the list
3. in the repository section of home, add the repository as physical disk
4. "present" (green up and down arrows) the physical disk on the vmserver itself (dont ask me why you have to do this but if you dont it wont find its own disk)

Repository problems

I have xml docs stored in the purchaseOrder table which I transferred there using xdb repository folders
via ftp.
I truncate table purchaseOrder.
Now I don't see any row in purchaseOrder. But I still see the repository showing the xml doc in
windows explorer(though not the contents).
(1)Whats is a repository. Is it only metadata and the contents are in all cases stored in Database tables.
(2)I am not able to remove the repository paths/links via any method
(sql error is given below, while i get similar error in win explorer and oracle en. manager).
SQL> delete from resource_view where any_path='/home/SCOTT/purchaseOrders/1999';
delete from resource_view where any_path='/home/SCOTT/purchaseOrders/1999'
ERROR at line 1:
ORA-31110: Action failed as resource /home/SCOTT/purchaseOrders/1999 is locked
by name
ORA-06512: at "XDB.XDB_RVTRIG_PKG", line 0
ORA-06512: at "XDB.XDB_RV_TRIG", line 9
ORA-04088: error during execution of trigger 'XDB.XDB_RV_TRIG'
(3)When I try to remove a resource from the repository for which corresponding content
exists in the database then I have no problem.

A resource is the general term (from the IETF WebDAV spec) that defines an object in a file / folder heirarchy. WebDAV defines a set of verbs for creating, manipulating and deleting resources, a set of meta data that a Dav server will maintain about each resource, and a communication protocl based on HTTP and XML for exchanging information between a WebDAV client and server.
In the case of XML DB for non-Schema based documents the content and the meta data are both stored in tables owned by the user XDB schema. For Schema based content the XDB tables maintain the meta data and the content is stored in the default tables defined / created when the schema was registered. The location of these tables is, by default in the relational schema of the user who registered the tables.
Resources are automatically created when WebDav or FTP clients send requests to the XML DB protocol handlers. THey can also be created via functions on the dbms_xdb PL/SQL package.
A Resource container is simple a 'folder' or 'directory'. Again these can be created via WebDAV or FTP. They an also be created using dbms_xdb.createFolder()...
I am not sure about what you mean by "In the specify a file at the URL option what value do we give...". Can you be more specific.
I anwered the question about partitioning by folder location in your other post on this topic..... It's a possible future direction...

Repository problem in Designer

I've install 8i DB in ora8i home and form 6i in default_home and after wards I install Designer 6i in default_home now when I Start Designer it gives error msg that 'the user does not have an installed repository run repository Admin Utility but on trying the same unable to Login with'system/manager' login as 'scott/tiger' and try to install repository but gives error 'CDR-21244 Process has been aborted. Causes : the previous process has failed or the user has aborted it.
Pls guide me how solve the repository re installetion. (also installed designer 2nd time but the problem continue)

It is very important to read the Online Documentation of Designer/2000 ( 6i) before Installing a Repository. There are steps like System Requirements etc and step by step about Installing the Repository. There are steps like New Tablespace, User , Roles etc and once must read these steps before attempting to Install the Repository. I have done this and it worked.
Hope this helps.
Thiru

IDM 6.0SP1 & Oracle as a repository problems

We use IDM 6.0SP1 + WebSphere 6 + Oracle 9 (+ oracle JDBC driver v10.x) as a repository with about 1 600 000 IDM accounts.
Each IDM account has at most 2 resource accounts.
We're facing two problems:
1) IDM is used to read changes (user creations and modifications) in 10 different DB2 tables (thanks to 10 DB2 active sync adapters) and provision a single LDAP directory based on those changes. We've about 10 changes per second to consume. The Oracle repository is about 32 Gigs .
We sometimes restart IDM but the adapters which are supposed to start automatically don't seem to start, or they start so slowly they look frozen.
We suspect bad Oracle response times and also we suspect each adapter triggers a full (Oracle) database scan when starting, which may take a while, in spite we pass a statistics script every night on Oracle.
We've already applied all the suggested/documented repository optimizations, so we wonder what else could we possibly do to improve IDM's interaction with its repository ? For example, is there any thing we can tune with the IDM RepositoryConfiguration XML object ?
2) In case of "long" (> 5 minutes) repository or network outage (provided that IDM and its repository reside on different servers), we noticed IDM adapters don't restart well automatically while configured to or they look frozen. We have to manually "try" different things in order for the adapters to start.
Most of commercial software relying on network or databases automatically deal with such outages so that they recover automatically, or at least, they just loose unsaved stuffs but they can restart anyway.
Is there such a feature with IDM (6.0SP1 or later) ? If not, what's the recommanded actions to take in order for the adapters to start and process the remaining
DB2 changes ?

IDM does some cleanup in its database when it starts. Perhaps that is what is slowing you down.
How many rows do you have in your "TASK" table?
Edited by: PaulHilchey on Mar 6, 2008 6:15 PM

Re: ActiveX and Repository Problems

Hi Jason,
Following is the Forte flag that I've used to redirect stdout of a
partition:
-fl c:\forte\log.txt(*:user1:1-63:255)
The rule of thumb that I know in case of segmentation violation is:
'Don't fight it, by-pass it!'
Hope this help a little.
Patrick Leveque
Reply-To: "Jason Carpenter" <[email protected]>
Our setup:
Clients: NT 4, Forte 3.0.F.2
Server: Unix, Forte 3.0.F.2
Central repository is located on the server, we are using ATTACHED shadows.
Our clients employ the use of an ActiveX control that is a spreadsheet-like
grid. As far as we know, it is the only spreadsheet-like grid that is
compatible with Forte. The ActiveX control is Protoview's Datatable v4.0.
Some of the time it behaves well in Forte. Many times it does not.
Situation:
While in debug mode (or after having pressed the running man) when we
attempt to interact with the grid, we get a Fatal Error on the task that
created the ActiveX control. Unfortunately this Fatal Error completely shuts
down our Forte clients. Usually there is not even time to look at stdout to
determine what the problem or error is. We have started sending stdout to a
separate log file, unfortunately the flags for logging are not documented
well enough for us to pick the right ones to determine what is really
happening.
In times past we have seen error messages to stdout that inform us that
there was a read/write error with the repository. Others mention
serialization/deserialization problems. All have a System Segmentation
violation error message. And we are always shut out of our client session.
>
We have noticed that sometimes when we integrate the problem will go away
for an hour or two. Other times integration does not help.
Question:
Has anyone else experienced this type of problem with ActiveX controls?
Does anyone have any suggestions to keep this problem from happening?
Jason Carpenter
CSC Consulting & Systems Integration
[email protected]
3811 Turtle Creek Blvd.
20th Floor
Dallas, TX 75219
214.520.0555
To unsubscribe, email '[email protected]' with
'unsubscribe forte-users' as the body of the message.
Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>
>
Get Your Private, Free Email at http://www.hotmail.com
To unsubscribe, email '[email protected]' with
'unsubscribe forte-users' as the body of the message.
Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>

Hi Jason,
Following is the Forte flag that I've used to redirect stdout of a
partition:
-fl c:\forte\log.txt(*:user1:1-63:255)
The rule of thumb that I know in case of segmentation violation is:
'Don't fight it, by-pass it!'
Hope this help a little.
Patrick Leveque
Reply-To: "Jason Carpenter" <[email protected]>
Our setup:
Clients: NT 4, Forte 3.0.F.2
Server: Unix, Forte 3.0.F.2
Central repository is located on the server, we are using ATTACHED shadows.
Our clients employ the use of an ActiveX control that is a spreadsheet-like
grid. As far as we know, it is the only spreadsheet-like grid that is
compatible with Forte. The ActiveX control is Protoview's Datatable v4.0.
Some of the time it behaves well in Forte. Many times it does not.
Situation:
While in debug mode (or after having pressed the running man) when we
attempt to interact with the grid, we get a Fatal Error on the task that
created the ActiveX control. Unfortunately this Fatal Error completely shuts
down our Forte clients. Usually there is not even time to look at stdout to
determine what the problem or error is. We have started sending stdout to a
separate log file, unfortunately the flags for logging are not documented
well enough for us to pick the right ones to determine what is really
happening.
In times past we have seen error messages to stdout that inform us that
there was a read/write error with the repository. Others mention
serialization/deserialization problems. All have a System Segmentation
violation error message. And we are always shut out of our client session.
>
We have noticed that sometimes when we integrate the problem will go away
for an hour or two. Other times integration does not help.
Question:
Has anyone else experienced this type of problem with ActiveX controls?
Does anyone have any suggestions to keep this problem from happening?
Jason Carpenter
CSC Consulting & Systems Integration
[email protected]
3811 Turtle Creek Blvd.
20th Floor
Dallas, TX 75219
214.520.0555
To unsubscribe, email '[email protected]' with
'unsubscribe forte-users' as the body of the message.
Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>
>
Get Your Private, Free Email at http://www.hotmail.com
To unsubscribe, email '[email protected]' with
'unsubscribe forte-users' as the body of the message.
Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>

Display Portlet Repository Problem - Page cannot be found

hi,
Just upgraded to 3.0.9.
when i go to Portal Homepage ..... Adminster..... Display Portlet Repository, and then select an item, i get the following error msg on my browser:
"The Page cannot be found"
Kin
null

Hi
The way I fix this problem was to Edit the portal provider and then click on Refresh Provider. The next time you go into Display Portlet Repository the item should be OK.
The refresh remote provider information - seems not to work.
Hope this helps
Kin

Pacman repository problems

So I noticed that pacman 2.6.1 allows for HTTP servers.
This is great, as I wanted to run my own repository (for personal packages).
Well I stuck the address in the pacman.conf as: Server = http://www.freewebs.com/frink36
and it says it fails to download the appropriate db.tar.gz file. I know its there, as I can download it from other places.
Is there something special I have to do? Ive tried removing the "www.".
Also, I dont have any of the actual packages uploaded (I wanted to test first), do I have to have all the packages uploaded for it to work?

nevermind, It was a problem w/ freewebs.com
Anyone have a good free place to host that has lots of space and no file upload limit?

Portlet repository problem

hello
i have suceesfully created a provider and its relative portlets. i can edit the provider with portal, but i can't see it in the portlet repository even when refreshing it.
does anybody knows why?
thanks

I have the same problem
I am using portal 3.0.6 on windows NT
I added the protlet provider successfully
but can not see it in portlet repository
even if I refreshed the portlet repository
please help !!
null

Mounting Repository Problem

Hi All,
When i mount MDM server a new Pop up windows open. In this Pop up window it shows Error mounting Server and File failed a CRC check. While In status it shows Communication Error. If anybody knows how to correct it Please do needful as soon as possible.
Thanks in advance,
Mandeep Saini

Hi Mandeep,
I fully agree with Cleopantra. I have faced this problem many times. this comes when you try to mount the MDM server using MDM console of different version as of MDM server.
To Resolve this, you need to install the MDM console of the same version as of MDM server.
Please check the MDM server version and install the console of the same version and your problem will be automatically solved.
To check the version of MDM server, open any log file. You will se following entries at the top of the log file"
<b>
Host name: MDM Server Name
Process id:
Compile type: RELEASE_32
MDM Server version: 5.5.41.70 Built on 2007-Jun-19
Repository schema version: 1.264
</b>
the statement MDM server Version tells the version of MDM server.
Hope this will help you. Revert if you have any more queries.
Thanks,
<b>Shiv Prashant Dixit</b>

Jolt repository problems

I am using wls 5.1, tuxedo 6.4 and Jolt 1.1. Does anyone know about a max number
of parameters for a particular service. Loading the jolt buffer seems to quit
after a service has approximately 270 parameters...unfortunately i need about
400 parameters. Does anyone know if the number of parameters can be configured
somewhere or if upgrading to a newer version of jolt will help? thanks

thanks for the advice, it has been very painful.
"Anthony Fryer" <[email protected]> wrote:
>
Mike,
Save yourself alot of pain and ditch the RE applet and use the jolt repository
bulk
loader instead.
Anthony
"Mike" <[email protected]> wrote:
Peter,
Thanks for the reply. I am running the applet from the same machine. Ihave
tried
both from the file and running it through the web server. I am wondering
if I am
not entering the correct values into the server and port fields. I have
been entering
"simpserv" and "3096" respectively. Is that correct? I have attached acopy
of my
ubb file.
Mike
Peter Holditch <[email protected]> wrote:
Mike,
Since the RE is an applet, it had to adhere to the sandbox restrictions.
Specifically, it can't connect to a remote machine, unless the remote
machine
is
also running an HTTP server that served the applet.
Either run it using appletviewer on the machine where the JSL is running,
or
start a webserver on the JSL's machine and run the RE in a browser viathe
webserver.
This is what I have found to be the problem in 90% of such cases...
Regards,
Peter.
Mike wrote:
I am trying to use the BEA Jolt Repository Editor. However when I startthe applet,
the user role, application password, user name, and user password fieldsare all
disabled. I have verified the the Jolt listener is running. What elsecould be the
problem?
thanks

File System Repository problem

Hello guys,
I'm on:
J2EE Engine 6.40 PatchLevel 98256.313
Portal 6.0.14.0.0
KnowledgeManagementCollaboration 6.0.14.4.0 (NW04 SPS14 Patch 4)
running on a Windows machine.
I'm trying to set up a File System Repository to connect to a remote share on another Windows machine. I followed the documentation and also some hints here from SDN and still can't connect to the share.
In component monitor it says:
Startup Error: The localroot does not exist:
10.80.1.55\Test_portal
In a trace log I'm getting this:
#1.5#000BCD3F823F00260000000800000AB000040D879131A728#1140775814806#com.sapportals.wcm.service.fsmount.FSMountService#sap.com/irj#com.sapportals.wcm.service.fsmount.FSMountService#Guest#59####3a9c7e20a51d11da928e000bcd3f823f#Thread[ConfigurationEventDispatcher,5,SAPEngine_Application_Thread[impl:3]_Group]##0#0#Error##Plain###Error mounting drive ,path
10.80.1.55
Test_portal#
#1.5#000BCD3F823F00260000000900000AB000040D879131ABAB#1140775814806#com.sapportals.wcm.service.fsmount.FSMountService#sap.com/irj#com.sapportals.wcm.service.fsmount.FSMountService#Guest#59####3a9c7e20a51d11da928e000bcd3f823f#Thread[ConfigurationEventDispatcher,5,SAPEngine_Application_Thread[impl:3]_Group]##0#0#Warning##Plain###FSMountService Reconfiguration (New path) failed to mount network path
pr1141
Test_portal - com.sapportals.wcm.service.fsmount.RemoteAccessException: Network component not started:
10.80.1.55
Test_portal
     at com.sapportals.wcm.service.fsmount.FSMountService.mountDrive(FSMountService.java:90)
     at com.sapportals.wcm.service.fsmount.FSMountService.updateConfigurables(FSMountService.java:349)
     at com.sapportals.wcm.service.fsmount.FSMountService.reconfigure(FSMountService.java:287)
     at com.sapportals.wcm.crt.CrtThreadSafeComponentHandler.tryToReconfigure(CrtThreadSafeComponentHandler.java:297)
     at com.sapportals.wcm.crt.CrtThreadSafeComponentHandler.handleReconfigure(CrtThreadSafeComponentHandler.java:147)
     at com.sapportals.wcm.crt.CrtComponentManager.reconfigureComponent(CrtComponentManager.java:343)
     at com.sapportals.wcm.crt.CrtConfigurationEventListener.notifyComponentConfigChanged(CrtConfigurationEventListener.java:79)
     at com.sapportals.wcm.repository.runtime.CmConfigurationProvider.sendChangeEvent(CmConfigurationProvider.java:1288)
     at com.sapportals.wcm.repository.runtime.CmConfigurationProvider.sendChangeEventOrRestart(CmConfigurationProvider.java:1237)
     at com.sapportals.wcm.repository.runtime.CmConfigurationProvider.configEvent(CmConfigurationProvider.java:256)
     at com.sapportals.config.event.ConfigEventService.dispatchEvent(ConfigEventService.java:210)
     at com.sapportals.config.event.ConfigEventService.configEvent(ConfigEventService.java:98)
     at com.sapportals.config.event.ConfigEventDispatcher.callConfigListeners(ConfigEventDispatcher.java:299)
     at com.sapportals.config.event.ConfigEventDispatcher.flushEvents(ConfigEventDispatcher.java:248)
     at com.sapportals.config.event.ConfigEventDispatcher.run(ConfigEventDispatcher.java:105)
I had an idea that it could be because of the Serverlet Engine User, which is set up due to documentation, but is only a local user not a domain user. I don't know to which extent this user is important for FS Mount Service. I thought, that for connecting to a remote share, there is only important the user and password set up in Network Path settings in CM.
Changing the user on which is EP running is quite a difficult step due to our security policies and so on. So I would rather not experiment with this until I try all the other possibilities.
Could you, please, give me some advice?

Hi Jiri,
have you used double forward slashes in your network path? Please try: //10.80.1.55/Test_portal
You can also test from the portal machine that you can mount the drive via command shell:
"net use t:
10.80.1.55\Test_portal /user:DOMAIN\USER"
(make sure you specify the user with domain)
Hope this helps,
Robert

Content Repository problem.

Hi All,
1) In OAC0 when I try to send the certificates it gives me the folowing error "Error in HTTP Access: IF_HTTP_CLIENT->RECEIVE 1". The test connection was successful.
2) Is there a way I can hide other std storage categories like DMS_C1_ST and just display the one I created?
Thanks
Paddy.

Hi,
There is way to hide other repository is just to delete those standard one given by same.
But the better solution is to default the repository for the users.
In IMG , create a profile. Assign the roles or users to profile. Assign the file applications ,storage CAT to this profile. In this way the need to user to get storage CAT will be fulfilled.
Your question if came from user point of view the above is best solution as users will get confused in Production environment if they see all category and by mistake they make pick up wrong category in hurry.
Regards,
Hemant

Unit Test Repository Problems

Ciao,
This is related to SQL Developer 3.0 Unit Test feature. Is it possible to have many shared repository in a single DB instance? How to prevent the users to delete the testcases written by another user in a shared repository?
Regards,
Bala

Hi, it is possible:
create the two roles
UT_REPO_USER
UT_REPO_ADMINISTRATOR
Extract the file dbtools\unit_test\repository\create10.sql from the jar oracle.sqldeveloper.unit_test.jar
Run the create10.sql script
HTH
Oliver

CM Repository problem

Similar Messages

Maybe you are looking for