C5 is a simple, reliable, and scalable open source database which improves on HBase in every way. It is optimized for fast failover and can be used in production for both OLTP and analytics, eliminating a whole class of pipelines and bottlenecks. It is the first HBase successor, fully HBase API-compatible, developed from the ground up to run on the cloud installs with very easy to no maintenance and tuning. Failover happens instantly, which makes it a viable option for APIs, and all the data is accessible to Hadoop analytics flows right away. C5 takes full advantage of high speed IO (SSDs) and is simple to grow to hundreds and thousands of nodes as needed.
A lot less than you'd think - the fundamentals of HBase are excellent. File formats, compacting, etc is all solid and the same.
The issues arise during fail over and the architecture of how HBase and HDFS interact. That is where the most changes are, and while it seems scary, I think people will find the changes most refreshing, and cut out literally hundreds of thousands of lines of code out of the write and read paths!
That's a great way to think about it. We want to push it back into the HBase ecosystem as well. Users can use tools like presto and impala to access the data via sql.
- We say we can do failover in a couple of seconds. We want to make it subsecond, but we can't do that reliably yet. In HBase this story is much more mixed.
- We wanted to really reduce complexity, as a result, you can just apt-get install c5 on each node and you are done. It's one daemon, one log file, and that's it. No xmx nonsense, and almost no tuning or config files. I don't know if you have dealt with hadoop before, but the complexity is high.
- Finally we have a much more advanced wireformat. In fact it's advanced by being simple (protobufs + http). As a result clients in languages other than java become very easy, without a thrift client.
In the future expect even cooler features, but for v1 we are just aiming for HBase compat.
I should also mention we had to work on our own version of protobufs. It's totally compatible but much faster in java. Check our blog for details. It's already open source!
I think he got the same question that i , we want to test the database and begin to do fun stuff with that but we dont want to call or write email (not that i a problem with that), is just curiosity if going to have a botton "sing in" or "try" ?
Yeah, I was really interested in messing around with it, but was being lazy and would have rather just entered my credit card instead of having to call or email.
We benefited greatly from Hadoop and it's legacy. Hadoop solves hard problems for some of the biggest companies. On the other hand it is very complicated to install and manage. Additionally HBase still has a ways to go with bindings in other languages and availability. By fixing this it should expand the set of use cases that Hadoop can be applied to.
We think we have built an enhanced version of the bottom of Hadoop that is a pleasure to run. We hope you find it pleasurable as well.
46 comments
[ 3.5 ms ] story [ 110 ms ] threadThe issues arise during fail over and the architecture of how HBase and HDFS interact. That is where the most changes are, and while it seems scary, I think people will find the changes most refreshing, and cut out literally hundreds of thousands of lines of code out of the write and read paths!
- We say we can do failover in a couple of seconds. We want to make it subsecond, but we can't do that reliably yet. In HBase this story is much more mixed. - We wanted to really reduce complexity, as a result, you can just apt-get install c5 on each node and you are done. It's one daemon, one log file, and that's it. No xmx nonsense, and almost no tuning or config files. I don't know if you have dealt with hadoop before, but the complexity is high. - Finally we have a much more advanced wireformat. In fact it's advanced by being simple (protobufs + http). As a result clients in languages other than java become very easy, without a thrift client.
In the future expect even cooler features, but for v1 we are just aiming for HBase compat.
We think we have built an enhanced version of the bottom of Hadoop that is a pleasure to run. We hope you find it pleasurable as well.
> C5 is a next generation distributed NoSQL system.
Is it called OhmData or C5?