# FoundationDB vs. other alternatives

**URL:** <https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989>\
**Category:** Using FoundationDB\
**Created:** [January 2, 2019, 11:26pm UTC](https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989 "2019-01-02T23:26:29Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![JohnZ622](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/johnz622/32/467_2.png) [@JohnZ622](https://forums.foundationdb.org/u/JohnZ622)\
**Post date:** [January 2, 2019, 11:26pm UTC](https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989/1 "2019-01-02T23:26:30Z")

</div>

Hi,

Can someone comment on the difference between FoundationDB and various NewSQL alternatives such as:  
Oracle RAC (might not be NewSQL since it is shared-everything)  
Clustrix  
MemSQL Cluster

Thanks,

---

<div class="post-metadata">

**Author:** ![alexmiller](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/alexmiller/32/326_2.png) [@alexmiller](https://forums.foundationdb.org/u/alexmiller)\
**Post date:** [January 3, 2019, 10:19pm UTC](https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989/2 "2019-01-03T22:19:27Z")

</div>

Well, largely, FoundationDB isn’t NewSQL. As far as I’m aware of how these terms are used, NewSQL requires there being a way to speak SQL to the database. Maybe someday if/when a SQL layer exists for FDB, it can graduate from NoSQL to NewSQL.

If you’re fine without SQL, or are content with something like the [document layer](https://www.foundationdb.org/blog/announcing-document-layer/), then as far as I’m personally aware, the selling point of Oracle RAC and Clustrix is that they try to be like your previous single-machine Oracle/MySQL setup, but can scale out. FoundationDB would fit closer to these two products, as it aims to be a scaleable OLTP database. I’ve generally heard of MemSQL more in the OLAP context, which currently isn’t FoundationDB’s strong suit due to the 5 second transaction limit. (Though there’s work going on to change that for read-only transactions.)

---

<div class="post-metadata">

**Author:** ![JohnZ622](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/johnz622/32/467_2.png) [@JohnZ622](https://forums.foundationdb.org/u/JohnZ622)\
**Post date:** [January 15, 2019, 6:07am UTC](https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989/3 "2019-01-15T06:07:58Z")

</div>

How much would it be to write a SQL layer? I don’t know the details too much but it is “just” the query engine right? 🙂

---

<div class="post-metadata">

**Author:** ![alexmiller](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/alexmiller/32/326_2.png) [@alexmiller](https://forums.foundationdb.org/u/alexmiller)\
**Post date:** [January 16, 2019, 1:12am UTC](https://forums.foundationdb.org/t/foundationdb-vs-other-alternatives/989/4 "2019-01-16T01:12:38Z")

</div>

PingCAP’s [TiDB](https://github.com/pingcap/tidb) would be reasonable similar to a SQL layer for FoundationDB. TiDB is sitting at almost 200k lines of code right now, so probably not a trivial undertaking.