Series of quaterly newsletters published by Data Virtualization Customer Support Team.
SHAPING THE FUTURE
Welcome to the sixth edition of our quarterly Customer Support Newsletter! In this issue you will find the latest news and many useful tips to help you get the most out of your TIBCO Data Virtualization (TDV) usage. Below is a summary of how our Data Virtualization Global Support team worked with customers and partners to resolve various TDV product and operational issues.
Q1'FY18 Customer Support Highlights
Customer Satisfaction (CSAT) = 9.4 (out of 10)
Mean Time to Resolution (MTTR) = 20 days
Total Support Tickets Closed = 250
Thank you for being part of our growth. Your success is our top priority and our goal is to help you achieve it. We look forward to many years of working together.
Please feel free to send your feedback to me at email@example.com
Sumit Chopra, Head of Customer Support
CUSTOMER SUCCESS STORY
An investment firm reported cluster instability and performance Issues experienced intermittently on fairly new version of CIS
Remove the nodes from the cluster mode and operate as stand-alone.
Observe if you still experience the issue.
If yes, its most likely a non-cluster issue that leads to the performance issues. Customer still experienced performance issues in stand-alone mode.
Gathered thread dumps, logs and heap dumps.
From the logs, it was pointed that there were a few culprit queries that were using the maximum allowed memory per request.
Increased the heap size and fine-tuned the memory settings, but they were still having server hung situations.
Pointed out the multi-table caching errors, but the customer dismissed those errors as red herring.
Examined the query plans of the culprit queries, but everything seemed to push fine.
From the logs, the IN clause appeared to be huge.
Despite being pushed and not giving any errors, it can drain the memory usage of a query.
When all the memory quota is consumed for a query, instead of failing, it starts to spill on disk.
This makes the execution extremely sluggish.
From the thread dumps, there were Oracle queries that were getting stuck at the database.
At each incident that the customer reported, we identified different set of problems.
In some incidents, some of the Cognos queries were not pushing.
Known repository bug regarding the auto-vacuum was occurring.
DBChannel queue size was adjusted for optimized performance.
Fetch size for bytes was adjusted for faster data fetching.
All of this work was done in several sessions with the customer, after multiple occurrences happened.
Gradually, customer started to educate their developers about the distinction of an "efficient query" versus an 'inefficient query'.
All the performance tuning and this key distinction helped the customer stabilize their environment.
Customer sentiments were probably tidal during the whole resolution process. It required deep dive collaboration and research from the customer and vendor's end.
Eventually, with a combination of using the best development practices, fine tuning the memory settings, server level configurations and resolving known (already fixed in newer TDV releases) repository issues, we were able to resolve the performance related issues and achieve a success story!
New TDV Ideas Portal
Check our the new TIBCO Ideas Portal. Customers can now submit their product feature enhancements directly to Product Management.
Recent Hot Fixes
7.0.6 HF1 released on February 6,2018
Download Hotfix (Login Required)
Next Hotfix - 7.0.6 HF2, scheduled to release on April 15, 2018
Note: If you have any issues downloading TDV Software from the Support Portal (eDelivery site) please contact Download Admin
Vacuuming the Postgres repository db inside TDV
In TDV versions 7.0.0 through 7.0.6, the Postgres meta db (aka the repository) can grow excessively on disk.
For details, please see the Knowledge Base link below.
This scenario can manifest in these ways:
- unplanned disk space usage above the typical upper range of ~1 GB.
- sluggish response seen in the Studio, for any UI action.
- potentially contribute to (but not the sole cause of) frequent unexplained cluster node shunning.
We recommend that a periodic inspection of the overall disk space usage or of the individual tables become a regular maintenance task and if needed, a manual cleanup be performed as described in KB 000035104
The future TDV v7.0.7 (estimated to release in May 2018) contains various fixes and improvements that will allow automatic vacuuming of the Postgres db to succeed, making this maintenance task less needed by you.
Knowledgebase Articles of Interest
Memory Sizing for your TIBCO Data Virtualization Server
TDV Join Algorithm and how to optimize joins
Autovacuum: Repository log is getting filled with autovacuum messages / misconfiguration of Postgres databases.
My client (JDBC or ODBC) on Windows has stopped connecting to my TDV server (on Linux)
"Uma helped us in a very timely manner with the best advise. I would definitely like him to be assigned to our future cases. Greatly appreciated his knowledge and help on this case." - Cable Company
"Always a pleasure interacting with Neelesh over CIS queries/support. Goes out of of the way to ensure a premium customer satisfaction experience." - Networking Company
"Ajay did an outstanding job assisting us through this issue." - Financial Company
"Vinod was very responsive to my questions, sought timely clarification from engineering and provided me with the answer I was looking for in a reasonable time frame." - Oil Company
"As always, Girish was able to quickly resolve the issue we were facing for couple of weeks. This even helped clear another issue which we were thinking was unrelated." -Pharmaceutical Company
TIBCO Software Inc., 3307 Hillview ave, Palo Alto, CA 94304
Forward this email | Update Profile | About our service provider
Sent by firstname.lastname@example.org