Cassandra at scale
Overall Satisfaction with Cassandra
It’s one of the database platforms we offer to the development community in our organization. We have various selections when it comes to databases including DB2, SQL Server, Oracle, and hadoop for data warehousing. Cassandra becomes the choice when developers want to use a highly available NoSQL db.
Pros
- Availability
- Fast performance
- Horizontal scalability
- Memory first
- Partition based
Cons
- Dealing with tombstone
- Maintenance/upgrade
- Compaction and repair
- High availability, which makes us use the tool
- Partition key distribution mechanism is well defined
- Horizontal scalability when needed for high volume days
- Complex tombstone management
It was packaged with the vendor product we bought. Also, it’s good for high performance transactional systems. I'm part of our NoSQL team and Cassandra quickly became a favorite for developers with agile teams.
Cassandra Feature Ratings
Using Cassandra
100 - Uuid lookup for multiple line of business.
Session and cookies management.
Records tracking system.
Session and cookies management.
Records tracking system.
100 - Java developers, operational suppor team with DBA background. Skills needed to support are dustribued system knowledge, os level administration skills, scripting is a plus and data modeling experiences.
- Document id lookup system
- Tracking if internal application usage
- Session management
- Fraud Risk system that determines any users might be potential victims
- Protecting internal uuid exposing to all other systems
- Document ID lookup system
- Url smart links
- Credit reports tracking system
- Asset management
Evaluating Cassandra and Competitors
Yes - Oracle and it was the decision from vendor
- Product Usability
- Product Reputation
- Prior Experience with the Product
I was not the decision maker in thus process however I like the easy use of OpsCenter .
Include more pilot groups ie developers to try out different features and do more stress testing.
Cassandra Implementation
- Implemented in-house
- Third-party professional services
CA and broadcom
Yes - Dev/test/prod and slow rollout when implementing in prod.
- Too many nodes to handle
- Upgrade can be difficult when you have too many clusters with too quick changes on versioning
Cassandra Support
Pros | Cons |
---|---|
Quick Resolution Kept well informed Immediate help available Support understands my problem Support cares about my success Quick Initial Response | Escalation required |
No - Our budget is not ready for this expense.
Yes - They recommended to upgrade to next version.
We have license with Datastax so have not dealt with apache.
Using Cassandra
Pros | Cons |
---|---|
None | None |
- Opscenter
- Bouncing servers
- Backup / restore
- Tombstones
- Repair process
- Compaction
Comments
Please log in to join the conversation