Discussion of Facebook’s data management technologies. Related subjects include:
At the highest level:
- Presto is, roughly speaking, Facebook’s replacement for Hive, at least for queries that are supposed to run at interactive speeds.
- Teradata is announcing support for Presto with a classic open source pricing model.
- Presto will also become, roughly speaking, Teradata’s replacement for Hive.
- Teradata’s Presto efforts are being conducted by the former Hadapt.
Now let’s make that all a little more precise.
Regarding Presto (and I got most of this from Teradata)::
- To a first approximation, Presto is just another way to write SQL queries against HDFS (Hadoop Distributed File System). However …
- … Presto queries other data stores too, such as various kinds of RDBMS, and federates query results.
- Facebook at various points in time created both Hive and now Presto.
- Facebook started the Presto project in 2012 and now has 10 engineers on it.
- Teradata has named 16 engineers – all from Hadapt – who will be contributing to Presto.
- Known serious users of Presto include Facebook, Netflix, Groupon and Airbnb. Airbnb likes Presto well enough to have 1/3 of its employees using it, via an Airbnb-developed tool called Airpal.
- Facebook is known to have a cluster cited at 300 petabytes and 4000 users where Presto is presumed to be a principal part of the workload.
Daniel Abadi said that Presto satisfies what he sees as some core architectural requirements for a modern parallel analytic RDBMS project: Read more
1. Continuing from last week’s HBase post, the Cloudera folks were fairly proud of HBase’s features for performance and scalability. Indeed, they suggested that use cases which were a good technical match for HBase were those that required fast random reads and writes with high concurrency and strict consistency. Some of the HBase architecture for query performance seems to be:
- Everything is stored in sorted files. (I didn’t probe as to what exactly the files were sorted on.)
- Files have indexes and optional Bloom filters.
- Files are marked with min/max field values and time stamp ranges, which helps with data skipping.
Notwithstanding that a couple of those features sound like they might help with analytic queries, the base expectation is that you’ll periodically massage your HBase data into a more analytically-oriented form. For example — I was talking with Cloudera after all — you could put it into Parquet.
2. The discussion of which kinds of data are originally put into HBase was a bit confusing.
- HBase is commonly used to receive machine-generated data. Everybody knows that.
- Cloudera drew a distinction between:
- Straightforward time series, which should probably just go into HDFS (Hadoop Distributed File System) rather than HBase.
- Data that is bucketed by entity, which likely should go into HBase. Examples of entities are specific users or devices.
- Cloudera also reminded me that OpenTSDB, a popular time series data store, runs over HBase.
OpenTSDB, by the way, likes to store detailed data and aggregates side-by-side, which resembles a pattern I discussed in my recent BI for NoSQL post.
3. HBase supports caching, tiered storage, and so on. Cloudera is pretty sure that it is publicly known (I presume from blog posts or conference talks) that: Read more
|Categories: Cloudera, eBay, Facebook, Hadoop, HBase, Market share and customer counts, NoSQL, Open source, Petabyte-scale data management, Specific users, Yahoo||4 Comments|
The pessimist thinks the glass is half-empty.
The optimist thinks the glass is half-full.
The engineer thinks the glass was poorly designed.
Most of what I wrote in Part 1 of this post was already true 15 years ago. But much gets added in the modern era, considering that:
- Clusters will have node hiccups more often than single nodes will. (Duh.)
- Networks are relatively slow even when uncongested, and furthermore congest unpredictably.
- In many applications, it’s OK to sacrifice even basic-seeming database functionality.
And so there’s been innovation in numerous cluster-related subjects, two of which are:
- Distributed query and update. When a database is distributed among many modes, how does a request access multiple nodes at once?
- Fault-tolerance in long-running jobs.When a job is expected to run on many nodes for a long time, how can it deal with failures or slowdowns, other than through the distressing alternatives:
- Start over from the beginning?
- Keep (a lot of) the whole cluster’s resources tied up, waiting for things to be set right?
Distributed database consistency
When a distributed database lives up to the same consistency standards as a single-node one, distributed query is straightforward. Performance may be an issue, however, which is why we have seen a lot of:
- Analytic RDBMS innovation.
- Short-request applications designed to avoid distributed joins.
- Short-request clustered RDBMS that don’t allow fully-general distributed joins in the first place.
But in workloads with low-latency writes, living up to those standards is hard. The 1980s approach to distributed writing was two-phase commit (2PC), which may be summarized as: Read more
|Categories: Clustering, CouchDB, Data warehousing, Databricks, Spark and BDAS, Facebook, Hadoop, MapReduce, Sybase, Theory and architecture, VoltDB and H-Store||1 Comment|
My California trip last week focused mainly on software — duh! — but I had some interesting hardware/storage/architecture discussions as well, especially in the areas of:
- Rack- or data-center-scale systems.
- The real or imagined demise of Moore’s Law.
I also got updated as to typical Hadoop hardware.
If systems are designed at the whole-rack level or higher, then there can be much more flexibility and efficiency in terms of mixing and connecting CPU, RAM and storage. The Google/Facebook/Amazon cool kids are widely understood to be following this approach, so others are naturally considering it as well. My most interesting of several mentions of that point was when I got the chance to talk with Berkeley computer architecture guru Dave Patterson, who’s working on plans for 100-petabyte/terabit-networking kinds of systems, for usage after 2020 or so. (If you’re interested, you might want to contact him; I’m sure he’d love more commercial sponsorship.)
One of Dave’s design assumptions is that Moore’s Law really will end soon (or at least greatly slow down), if by Moore’s Law you mean that every 18 months or so one can get twice as many transistors onto a chip of the same area and cost than one could before. However, while he thinks that applies to CPU and RAM, Dave thinks flash is an exception. I gathered that he thinks the power/heat reasons for Moore’s Law to end will be much harder to defeat than the other ones; note that flash, because of what it’s used for, has vastly less power running through it than CPU or RAM do.
|Categories: Amazon and its cloud, Buying processes, Cloudera, Facebook, Google, Intel, Memory-centric data management, Pricing, Solid-state memory||19 Comments|
There’s much confusion about Cloudera’s SQL plans and beliefs, and the company has mainly itself to blame. That said, here’s what I think is going on.
- Hive is good at some tasks and terrible at others.
- Hive is good at batch data transformation.
- Hive is bad at ad-hoc query, unless you really, really need Hive’s scale and low license cost. One example, per Eli Collins: Facebook has a 500 petabyte Hive warehouse, but jokes that on a good day an analyst can run 6 queries against it.
- Impala is meant to be good at what Hive is bad at – i.e., fast-response query. (Cloudera mentioned reliable 100 millisecond response times for at least one user.)
- Impala is also meant to be good at what Hive is good at, and will someday from Cloudera’s standpoint completely supersede Hive, but Cloudera is in no hurry for that day to arrive. Hive is more mature. Hive still has more SQL coverage than Impala. There’s a lot of legacy investment in Hive. Cloudera gets little business advantage if a customer sunsets Hive.
- Impala is already decent at some tasks analytic RDBMS are commonly used for. Cloudera insists that some queries run very quickly on Impala. I believe them.
- Impala is terrible at others, including some of the ones most closely associated with the concept of “data warehousing”. Data modeling is a big zero right now. Impala’s workload management, concurrency and all that are very immature.
- There are some use cases for which SQL-on-Hadoop blows away analytic RDBMS, for example ones involving data transformations – perhaps on multi-structured data – that are impractical in RDBMS.
And of course, as vendors so often do, Cloudera generally overrates both the relative maturity of Impala and the relative importance of the use cases in which its offerings – Impala or otherwise – shine.
- A survey of SQL/Hadoop integration (February, 2014)
- The cardinal rules of DBMS development (March, 2013)
|Categories: Cloudera, Data warehousing, Facebook, Hadoop, SQL/Hadoop integration, Workload management||4 Comments|
Cassandra’s reputation in many quarters is:
- World-leading in the geo-distribution feature.
- Impressively scalable.
- Hard to use.
This has led competitors to use, and get away with, sales claims along the lines of “Well, if you really need geo-distribution and can’t wait for us to catch up — which we soon will! — you should use Cassandra. But otherwise, there are better choices.”
My friends at DataStax, naturally, don’t think that’s quite fair. And so I invited them — specifically Billy Bosworth and Patrick McFadin — to educate me. Here are some highlights of that exercise.
DataStax and Cassandra have some very impressive accounts, which don’t necessarily revolve around geo-distribution. Netflix, probably the flagship Cassandra user — since Cassandra inventor Facebook adopted HBase instead — actually hasn’t been using the geo-distribution feature. Confidential accounts include:
- A petabyte or so of data at a very prominent company, geo-distributed, with 800+ nodes, in a kind of block storage use case.
- A messaging application at a very prominent company, anticipated to grow to multiple data centers and a petabyte of so of data, across 1000s of nodes.
- A 300 terabyte single-data-center telecom account (which I can’t find on DataStax’s extensive customer list).
- A huge health records deal.
- A Fortune 10 company.
DataStax and Cassandra won’t necessarily win customer-brag wars versus MongoDB, Couchbase, or even HBase, but at least they’re strongly in the competition.
DataStax claims that simplicity is now a strength. There are two main parts to that surprising assertion. Read more
|Categories: Cassandra, Clustering, Couchbase, Data models and architecture, DataStax, Facebook, HBase, Health care, Log analysis, Market share and customer counts, MongoDB, NoSQL, Petabyte-scale data management, Specific users||10 Comments|
I chatted yesterday with the Hortonworks gang. The main subject was Hortonworks’ approach to SQL-on-Hadoop — commonly called Stinger — but at my request we cycled through a bunch of other topics as well. Company-specific notes include:
- Hortonworks founder J. Eric “Eric14″ Baldeschwieler is no longer at Hortonworks, although I imagine he stays closely in touch. What he’s doing next is unspecified, except by the general phrase “his own thing”. (Derrick Harris has more on Eric’s departure.)
- John Kreisa still is at Hortonworks, just not as marketing VP. Think instead of partnerships and projects.
- ~250 employees.
- ~70-75 subscription customers.
Our deployment and use case discussions were a little confused, because a key part of Hortonworks’ strategy is to support and encourage the idea of combining use cases and workloads on a single cluster. But I did hear:
- 10ish nodes for a typical starting cluster.
- 100ish nodes for a typical “data lake” committed adoption.
- Teradata UDA (Unified Data Architecture)* customers sometimes (typically?) jumping straight to a data lake scenario.
- A few users in the 10s of 1000s of nodes. (Obviously Yahoo is one.)
- HBase used in >50% of installations.
- Hive probably even more than that.
- Hortonworks is seeing a fair amount of interest in Windows Hadoop deployments.
*By the way — Teradata seems serious about pushing the UDA as a core message.
Ecosystem notes, in Hortonworks’ perception, included:
- Cloudera is obviously Hortonworks’ biggest distro competitor. Next is IBM, presumably in its blue-forever installed base. MapR is barely on the radar screen; Pivotal’s likely rise hasn’t yet hit sales reports.
- Hortonworks evidently sees a lot of MicroStrategy and Tableau, and some Platfora and Datameer, the latter two at around the same level of interest.
- Accumulo is a big deal in the Federal government, and has gotten a few health care wins as well. Its success is all about security. (Note: That’s all consistent with what I hear elsewhere.)
I also asked specifically about OpenStack. Hortonworks is a member of the OpenStack project, contributes nontrivially to Swift and other subprojects, and sees Rackspace as an important partner. But despite all that, I think strong Hadoop/OpenStack integration is something for the indefinite future.
Hortonworks’ views about Hadoop 2.0 start from the premise that its goal is to support running a multitude of workloads on a single cluster. (See, for example, what I previously posted about Tez and YARN.) Timing notes for Hadoop 2.0 include:
- It’s been in preview/release candidate/commercial beta mode for weeks.
- Q3 is the goal; H2 is the emphatic goal.
- Yahoo’s been in production with YARN >8 months, and has no MapReduce 1 clusters left. (Yahoo has >35,000 Hadoop nodes.)
- The last months of delays have been mainly about sprucing up various APIs and protocols, which may need to serve for a similar multi-year period as Hadoop 1’s have. But there also was some YARN stabilization into May.
Frankly, I think Cloudera’s earlier and necessarily incremental Hadoop 2 rollout was a better choice than Hortonworks’ later big bang, even though the core-mission aspect of Hadoop 2.0 is what was least ready. HDFS (Hadoop Distributed File System) performance, NameNode failover and so on were well worth having, and it’s more than a year between Cloudera starting supporting them and when Hortonworks is offering Hadoop 2.0.
Hortonworks’ approach to doing SQL-on-Hadoop can be summarized simply as “Make Hive into as good an analytic RDBMS as possible, all in open source”. Key elements include: Read more
Recently, I observed that Big Data terminology is seriously broken. It is reasonable to reduce the subject to two quasi-dimensions:
- Bigness — Volume, Velocity, size
- Structure — Variety, Variability, Complexity
- High-velocity “big data” problems are usually high-volume as well.*
- Variety, variability, and complexity all relate to the simply-structured/poly-structured distinction.
But the conflation should stop there.
*Low-volume/high-velocity problems are commonly referred to as “event processing” and/or “streaming”.
When people claim that bigness and structure are the same issue, they oversimplify into mush. So I think we need four pieces of terminology, reflective of a 2×2 matrix of possibilities. For want of better alternatives, my suggestions are:
- Relational big data is data of high volume that fits well into a relational DBMS.
- Multi-structured big data is data of high volume that doesn’t fit well into a relational DBMS. Alternative: Poly-structured big data.
- Conventional relational data is data of not-so-high volume that fits well into a relational DBMS. Alternatives: Ordinary/normal/smaller relational data.
- Smaller poly-structured data is data for which dynamic schema capabilities are important, but which doesn’t rise to “big data” volume.
|Categories: Cassandra, Data models and architecture, Data warehousing, Exadata, Facebook, Google, Hadoop, HBase, Log analysis, Market share and customer counts, MarkLogic, NewSQL, NoSQL, Oracle, Splunk, Yahoo||10 Comments|
The DataStax and Cassandra stories are somewhat confusing. Unfortunately, DataStax chose to clarify them in what has turned out to be a crazy news week. I’m going to use this post just to report on the status of the DataStax product line, without going into any analysis beyond that.
Once again, I’m working with an OLTP SaaS vendor client on the architecture for their next-generation system. Parameters include:
- 100s of gigabytes of data at first, growing to >1 terabyte over time.
- High peak loads.
- Public cloud portability (but they have private data centers they can use today).
- Simple database design — not a lot of tables, not a lot of columns, not a lot of joins, and everything can be distributed on the same customer_ID key.
- Stream the data to a data warehouse, that will grow to a few terabytes. (Keeping only one year of OLTP data online actually makes sense in this application, but of course everything should go into the DW.)
So I’m leaning to saying: Read more
|Categories: Analytic technologies, Cloud computing, Clustering, Data warehousing, dbShards and CodeFutures, Facebook, Infobright, MySQL, OLTP, Open source, Parallelization, Software as a Service (SaaS), Solid-state memory||13 Comments|