Content Feed

 Connections  WebSphere  IBM Connections 5 July 2016

Startup Beans and adding additional nodes in Connections

Last week, working with a customer that had setup a two-node install of Connections 5.5. Getting ready to go live and we noticed an issue in the SystemOut log of the InfraCluster when the server would start or when the search index task would run.

CLFRW0454E: Unable to get indexing information about the current server, the index will not be built for this server.

We had followed the instructions for adding the new node, and couldn't determine why we were seeing this issue. Opened a PMR and we got back, after the initial MustGather request, some information to try. Basically, the internal technote indicated that the Startup Beans Service wasn't enabled in the newly added App Server.  Going to the app server's Container Settings > Container Services > Startup Beans service page and enabling that option followed by a sync and restart solved the issue.


 Connections 20 June 2016

IBM Connections 5.5 CR1 and Connections Mail - issue with calendar picker

After updating a test server with Connections 5.5 CR1, a customer noticed that the calendar picker in Activities would stay open after selecting a Due Date for an activity. This didn't happen in 5.5 or earlier versions. We opened a PMR with IBM and after some back and forth, according to IBM the issue is related to Connections Mail being enabled on the server.  If you don't have Connections Mail enabled, there's no issue.

A fix is in the works.  I'm hoping that it will be an update available from FIx Central and not just a fix, but we'll see.

UPDATE July 1: My client received a fix and installed it and it has resolved the issue. LO89462 is the APAR number, so open a PMR with that info if you need the fix.


 Connections  WebSphere  SSL 10 June 2016

IBM Cnx 5.5 CR1 and WAS 8.5.5.8 update - something to remember

I was updating a Connections 5.5 environment to CR1 today and part of the process is updating WebSphere to 8.5.5.8.  The installs all went very smoothly untill we went to verify the updates by launching the site.  We saw the always generic "We are unable to process your request" message hitting the homepage.  Off to the SystemOut.log!

Looking at the logs, I saw a series of SSL Handshake errors and then it hit me - the Java policy files to handle the large SSL key size must have been overwritten.  I need to remember that they will be overwritten during an update like this.

Checking the \java\jre\lib\security folder verified that the files had been replaced.  Once we replaced them with the correct version and restarted WebSphere we were able to login successfully.

Now to finish the testing!

UPDATE: Rainer asked where the updated policy files were downloaded from.  I used this IBM link to download them (login required) and chose the first option that included the j9 VM2.6 version.  Use the "java -version" command in the WAS Java folder to confirm that the JVM matches the versions listed there.


 Connections  IBM Docs  IBM Connections Docs 3 June 2016

IBM Docs - Mapping CIFS drives

Working on a Docs 2.0 install and came across some strange issues when setting up the servers to map drives to the Windows file share. On some servers, the drives mapped as expected, while others didn't.  No errors were logged in the Windows Event Viewer.

For reference, the documentation from IBM we followed is here.

To determine what was going on, I edited the cifsmount.bat to include some "logging" output to see if anything was happening we couldn't see when the task ran.  Oddly enough, I found messages like this:

X: has a remembered connection to \\fileserver\CnxData. Do you want to overwrite the remembered connection? (Y/N) [Y}:


So, WIndows was trying to do what we asked, it just had a mapping in its memory and needed a response.

To fix this I altered the cifsmount.bat file to 1- delete any mappings for the drives I'm using, and 2 - added a "/y" to the net use line so that it would default to Yes for any prompts.

net use /y X: \\fileserver\DocsData >>d:\logs\mountX.log


I kept the logging in so we can check if there are issues in the future.

Hope this helps someone else.


 Domino  Verse  Traveler 25 May 2016

Known Issues after upgrading to Verse 9.4 for Android and Something That Worked For Me

This evening the following technote was published by IBM:

Overview of known issues affecting users after upgrading to Verse 9.4 for Android

After I upgraded I was having issues syncing. One thing I noticed was that if I went into the main Verse app the prompt to force a new sync would appear, but the sync never finished. Each time I opened the Verse app, I saw the same prompt.

This morning I finally thought about deleting the local data and trying again. After doing this, I was prompted to login to my Traveler server and it performed a full sync and I've been working as expected since.

Hope this helps and sorry for the very long, IBM-like title.


 Connections  DB2 25 May 2016

DB2 HADR configuration reminder

This one is a reminder for me more than anything:

When setting up DB2 HADR, for IBM Connections, IBM Sametime, or anything, don't forget to make sure you've set the logging for the database before trying to configure the standby database.

It's easy to do when you're working with the number of databases that Connections involves and that's what happened here. We somehow skipped over one of them, even though we were using the DB2 directory as a guide. We were trying to start the database as a standby - one of the final steps - and got the SQL1767N error, reason code=1.  This indicates that the database wasn't in a roll-forward pending state which is what we want for setting up the standby.

After looking around I decided to check the log folder on the primary server and found that there weren't any folders for this database in the archive log folder.  So this meant that since the database wasn't configured for logging, it wouldn't be in the state we wanted after a restore.

This was a quick fix of just running the commands we needed to setup the database for logging on the primary side and then taking a new offline backup and restore it on the standby server and run the commands.  Lo and behold - it did what we expected - started as the HADR standby.

I should probably work on scripting this out to avoid this in the future.

Hope this helps someone else!


 Connections  DB2 12 January 2016

IBM Connections dbWizard issue - it’s all about the bits

Started a Connections 5.5 deployment for a customer last week. Things are going well, but when we got to the database Wizard, the wizard wan't cooperating. We couldn't choose the DB2 instance - which normally auto-populated anyway. We basically could bounce back between the Create/Update/Delete menu and the Database type and couldn't move forward.

We were able to create databases manually, but with other deployments coming up we wanted to make sure it was working.

After looking around the systems, I noticed some memory errors and discovered that the 32-bit version of DB2 was installed. Upgrading the version to 64-bit version resolved the dbWizard issue.

Just to be safe - and since we had only populated some test users - we dropped the original databases and created a new set using the Wizard.  

Moving on!


 Lotus Notes  Lotus Notes Domino 17 February 2015

Notes 9.0.1 FP3 IF1 - Updated Support for Google Calendar


Recently, the focus of the Domino community has been on the Interim Fixes and Fix Packs for the Domino server around the POODLE issues.  Yesterday I noticed a TechNote about an Interim Fix and reading through the Fix List for it, I can upon a fix for an issue we had noticed previously with a client.  That is that the federated calendar with Google was throwing an error.  The error was pretty generic, but it was something with authentication.

The Fix List yesterday referenced that they were updating the version of the Google Calendar API.  It appears that the API version that was in use was deprecated back in November 2014.  The new version prompts you to login to Google and accept that Notes will manage your calendar.  If you've used any other 3rd party product with Google, it will look familiar.  AND...it supports two-factor authentication as you'd expect.

Create the calendar as normal and you'll get a prompt like one above.

Federate away!

Interim Fixes for Notes 9.0.1.x


 Sametime 18 August 2013

Sametime Policy Naming - One character to not use in a name

Working with a client last week, they had reported an issue getting a policy setting from Sametime to apply properly.  The Chat History setting wasn't getting set as expected.

Looking at the issue, we determined that the policy file - policies.user.xml - wasn't getting updated. In fact, it had been a few days since it was updated. Looking at the trace output on the community server, everything looked OK. We could see the policy that was updated on the SSC getting read by the community server, and even see that the setting was what the client wanted. The XML wasn't getting updated.

Looking at it later in the week, I noticed a meeting policy had a date in it and that date coincided with the last date the policy XML file was updated. The date had a slash in it and I thought that was odd. Checking the documentation, it isn't listed as a prohibited characted, but I changed how the date was entered and waited to see if the policy XML file was updated. Lo and behold it was. Checking with a couple of users, they reported that the policy that was set incorrectly was now showing up as expected.

Two takeaways:

  • Do NOT use anything but alphanumerics in the policy name. Stay away from special characters.
  • Don't use dates in the policy name.
It would be good to have a comment field in the policy document that would allow some kind of tracking of changes, but that might need to be retained elsewhere for now. Also, I believe that changing the name of the policy whenever a change is made would potentially cause some issue with updating clients since they have to pick up what is essentially a "new" policy and then apply the "new" settings.

Coming this week - SPNEGO. Fun times ahead!


 Sametime 15 July 2013

Due by Thursday, July 18 - Update your Sametime Proxy certificates

The current set of security certificates used to send push notifications to Apple iOS devices expire this week.  So by the 18th, you'll want to download and install the "fix" for Sametime Proxy server that updates the certs.

See IBM Technote 1641787 for more info including instructions on installing the update.