<<<Be careful about assuming too much. I believe that Mongo only very recently got multi-document ACID functions, which are essential for any transaction which spans more than one document … which is most financial operations. It could be they now have this covered, but some careful analysis should go into evaluating this.>>>
Ummm. Yeah. It is also the response that a person defending the status quo from a disruptive technology takes. Of course it is not going to be able to do all that the well established and nearly perfected SQL databases can do. That is not the point.
The point is it can do it, it will get better at doing it, and more and more use cases for doing it will be found on the database.
That is what disruptive technologies do. And the fact that companies as smart as Baidu and Cisco and Microsoft are moving business critical functions to Mongo should demonstrate to all that this is the case.
Remember flash? What happened when Pure started rolling out flash? Oh yeah, a niche technology, never disrupt disks, best be careful where we put our mission critical data…almost put NTAP out of business.
It is also the response that a person defending the status quo from a disruptive technology takes.
On the contrary, I have been very interested in Mongo and other such databases for many years. I was introduced to it via the product formerly known as Data Direct Cloud, now Hybrid Data Pipeline which provides SQL and REST(OData) interfaces to a large number of cloud and on-prem data sources including major applications and a multiple database types including a number of NoSQL databases.
Nothing I have said is intended to be negative about Mongo as an investment. There are clearly a large number of use cases where it or something like it is a preferred solution. I am simply trying to express a bit of caution because some of the statements sounded as if they expected major migrations from relational DBs, which are going to be with us for a long time.
On October 19, 2017, MongoDB shares rose nearly 34 percent on their first day of trading in a $192 million IPO. The technology company listed its stock on the Nasdaq under the symbol “MDB.” MongoDB priced its 8 million-share initial public offering at $24 a share Wednesday night, above the expected range of $20 to $22 per share.
If my math is correct, the 180 days would expire on April 17, 2018.
The specs vs. each other. note, that this blog, is from May of 2017 and is therefore about a year behind. The issue with transactional work will soon (if all works out well) not be an issue as specified in this blog.
The blog, if you read through it, will tell you why the name “Mongo” was chosen.
The strengths, at the time that this blog was written. For Apache, need enormous database to scale fast.
For Mongo, utterly unstructured data such as from mobile, internet of things, and AI/ML.
Cassandra sounds like a great product.
Mongo sounds like a great product.
Of the two products, Mongo is going to add features and get better much faster.
It appears to me that Cassandra was made for the purpose of large data sets that SQL could not handle very well, and that it is very good at what it does and will maintain that focus.
But to review for you guys. I am going out tonight, so will not be digging further myself today. Perhaps others can pick up on this.
It’s hard to perform an apples to apples comparison given that open source databases like MongoDB and MySQL are predominantly deployed without commercial license. IDC and other analysts report of percentage of revenue shares but this fails to measure the huge number of open source database deployments.
Second problem is of deployment size. There’s fewer Cassandra deployments compared with MongoDB, but Cassandra implementations might be as much as an order of magnitude larger in terms of data storage. I’d love to know how much data was stored by each DBMS system but unfortunately we have now way to know that.
In this comparison https://db-engines.com/en/system/Cassandra;MongoDB which has a lot of details, in the vendor supplied information Cassandra claims to be the leading NoSQL database, but no metric is given to support this.
From that comparison, I have to say that Mongo sounds more technically interesting, but it is clear that Cassandra has very thoroughly addressed any scalability issues.
Several comparisons point to the third contender being HBase, which comes from the same source as Hadoop … but there are some suggestions that Hadoop is the actual choice for really unstructured big data.
For Mongo vs Cassandra there are number of sources to compare, but here are some inputs from stackoverflow (which is a resource for developers to answer developer questions):
Though Cassandra is NoSQL, it does not seem to offer the same flexibility than Mongo offers for dynamic schemas and hierarchical data. For developer productivity and faster development turnaround time, Mongo will win out. No wonder it is very popular among developers.
Cassandra ensures no single point of failure which ensures 100% availability. Probably that is Netflix opted for Cassandra. The list of companies using it is impressive too. Though it seems Mongo is steadily improving in ensuring all time avilability. I read somewhere max outage was 40 seconds, with a comment below correcting it that the lastes was only two seconds.
Of course there will be number of other differences. The key difference from a business stand point Cassandra is backed by Apache Foundation which I believe is a non profit with only free version. Mongo has a hybrid model with free and premium versions. Because of it, I believe there will be more resources spent on development of Mongo and ecosystem around it - think businesss intelligence connnectors, visualization tools, eventually industry specific vertical applications etc.
Overall Mongo is number 5 in ranking, the first 4 being relational database. Interestingly though Mongo is growing really fast, the database that grew faster last two years than Mongo was postgreSQL. The winning thing about postqresSQL is it is a RDBMS database with some flexibility to handle NoSQL. Having no transactions was a handicap for NoSQL limiting its outreach. It will be very interesting to see what happens with the new transactions capability in Mongo.
If I look at db-engines ranking Mongo is the number one NoSQL database by quite some distance.
Yes, but that ranking is based on number of citations on web sites and forums and the like, i.e., how many people are talking about it. It is not based on number of installations, total data under management, revenue, or any other measure which relates to real success in implementations.