 So students in this module I will of course throughout this last modules I have been comparing and talking about NoSQL vis-a-vis RDVMS model. Now in this module I will do very very specific and very focused comparison of NoSQL with RDVMS and that comparison will be based upon 10 criteria or you can say I will tell you the 10 major differences between the NoSQL and the RDVMS module. So let's look at those 10 differences and then in the subsequent modules I will briefly go over each of those 10 differences as a pair of differences. So on your screen you can see these 10 differences okay or the 10 advantages the word advantage is more important. The 10 advantages over here okay the 10 advantages not the differences of course we have discussed that NoSQL is not is not relational as opposed to MySQL which is relational and of course we will discuss it more also in some other course. So these are the advantages of NoSQL over RDVMS. So these 10 so now let's go and look at these 10 advantages in more detail. So let's ETL ETL is Extract Transform Load okay now this is the domain of data warehousing which is course 36 CS614 a lot of details over there but as over here as is data storage is used why because NoSQL supports these three formats these three formats are not the four types of data structures remember big difference so since these three formats are supported so there is less coding so there is less effort there is less plumbing code as the term goes that we convert the formats and then make them usable. So these three formats are supported so since these three formats are supported there is less ETL and there is less data conversion when there is less data conversion of course there is less maintenance and when there is less maintenance the maintenance cost goes down and it is multi language NoSQL and faceted search now what is faceted search faceted search does is that when you get the search results then certain filters are applied to narrow down the search okay and those filters come from you identify certain mutually exclusive but totally exhaustive okay strings or keywords based upon which this faceted search is performed. So this is very powerful feature for those scenarios in which we have kind of a text documents or document database which is supported by NoSQL of course we have this pattern matching in RDB mass relational but it is very slow over there is very slow and it is not it is not meant and built for this kind of searches then is the named entity extraction entity is something with our interest such as the name of a person name of a place a date name of an organization right and those things are supported those are part of the NoSQL solution which are not there in relational RDB mass model relational model this is a big advantage okay then is handles change over time okay and no SQL magic needed now as we have been discussing again and again that NoSQL is agnostic schema agnostic it means that when we start populating a NoSQL database you know it's not required to have a schema do you populate the database NoSQL database and work on it and it is when you need to know it when you perform a read on it that we have discussed when you perform a read on it then you need to know or what kind of schema should have been there so it is it is flexible the advantage is that as the schema changes over time in a relational model when you modify the schema it has a great ripple effect many things have to be changed many things have to be maintained it's a very expensive endeavor in terms of time and resources universal indexing immediate use as the NoSQL database is populated the indexes are also updated so it doesn't take a lot of time and you don't have to master and mastering SQL is a thorny art you use which type of joints outer join okay nested loops or select within select or cursors or sensitive cursors what else what more okay so it is a complex endeavor over here you don't need that kind of complexity to master to retrieve what you are looking for and SQL a NoSQL is for compatibility purposes they are different things NoSQL is different but of course it can support for the compatibility and it frees NoSQL SQL frees the developers from the vagracies from the complexities of the database so that they can spend more time on being more productive and of course the cost also NoSQL runs on inexpensive servers a lot of inexpensive servers which is horizontal scaling versus buying a single expensive server which is vertical scaling for a relational model and this this inexpensive hardware increases the what yes availability and adds durability because when you have multiple servers servers can crash this is a fact of life but when you have multiple servers horizontal scaling okay then is very less likelihood that all the servers fail maybe one can fail but in vertical scaling if the one server fail everything stops not over here and there are variety of NoSQL databases to meet your needs as opposed to a single which is the relational model range of data structures you can look at over here and there are wide variety of new entrants also such as IBM such as Microsoft such as Oracle it is not just limited to open source okay and finally there is no historical batch package there are no lots and lots of historical data that needs to be converted and you can perform the aggregations okay without porting lot of data so it is next to the data okay and indexes are updated as the as the databases are populated therefore it is this this aggregation is close turn around time is close next to the data so these are some of the fundamental basic 10 advantages of NoSQL with respect to the relational model that is all I have for this module for you