Numbers Every Database Professional Should Know

Latency Estimates in a Processor

In this post I will setup the next post by reminding you of these numbers every programmer should know. The picture shows the latency to access data across the three levels of cache in a modern processor, and across the memory bus to DRAM, and then across to an SSD or spinning disk drive. 

These numbers are the key to in-memory database performance. L1 cache is expensive and small. All data and all instructions are fed to the core through the L1 cache and typically the instruction and the data it will operate on are each fetched independently. If the data (or instruction… but I will just talk about the data from here on out) is not found in the small L1 cache the L2 cache is searched. Likewise, if the data is not in the L2 cache the L3 cache is searched and then DRAM. As you can see there is a 200X performance difference between a fetch from L1 cache and a fetch from DRAM. If the data must be fetched from SSD or disk, there is a 1000X penalty. 

Databases work to mitigate the 1000X penalty by pre-fetching data into DRAM. Modern processors mitigate the penalty between cache and DRAM by trying to anticipate the data required and pre-fetching it into L3 cache. This is a gamble and the likelihood of winning the pre-fetch bet varies with your DBMS and with the workload. More simultaneous queries against many tables puts pressure on the memory and lessens the effectiveness of data pre-fetch. 

Databases also mitigate the 1000X penalty by compressing data so that each fetch of a block from disk captures more data. For tabular (not columnar) data each block must be decompressed before it can be used. Decompression steals CPU cycles from query processing but mitigates some of the cost of I/O to disk. Columnar databases fetch compressed data and, if the databases use modern vector processing techniques, they operate directly on the compressed vector data without decompression. This is especially powerful as these vectorized columnar databases can push compressed data into cache and use that memory more effectively. 

Vectorized data can use the supercomputing instruction set that includes SIMD instructions. SIMD stands for Single Instruction; Multiple Data and these instructions fetch a single instruction from L1 cache and reuse it on the vectorized data over and over with pausing to fetch new instructions. Note that in a modern processor waiting for instructions or data to be fetched from cache or DRAM appears as CPU busy time. When the processor is stalled waiting for data to be fetched from disk the query/job gives up the CPU and dispatches a new query/job. 

Finally, for the longer-running queries associated with analytics and business intelligence (where long-running means a few seconds or more… “long” in CPU instruction time), it is highly likely that the L1 and L2 cache in each core will be flushed and new data will be fetched. It is even likely that all the data in L3 cache will be flushed. In this case most of the goodness associated with compression and vectorization is flushed with it. The vectors must be fetched again from L3 and DRAM. Further, note that when you run in a virtual machine or in the cloud you may think that you are the only user but other virtual machines other containers, or other serverless processes may be flushing you out of memory all the time. 

This is important. Until we swapped out a query for a new one it looked like in-memory vector processing might be 200X-1000X faster than a tabular query; but in an environment where tens or hundreds of queries are running concurrently the pressure on memory keeps flushing the cache and the advantage of vectorized query processing is reduced. Reduced, not eliminated. We would still expect the advantages of compressed data in cache and SIMD supercomputing instructions to provide much more than a 10X speedup. 

As mentioned, this post is a detailed review of material I covered years ago when HANA was introduced. In the next post I will add some new thinking. 

Here is more information on processor architecture and cache usage.

A Short DW DBMS Market History: HANA, Oracle, DB2, Netezza, Teradata, & Greenplum

Here is a quick review of tens years of data warehouse database competition… and a peek ahead…

Maybe ten years ago Netezza shook up the DW DBMS market with a parallel database machine that could compete with Teradata.

About six years ago Greenplum entered the market with a commodity-based product that was competitive… and then added column store to make it a price/performance winner.

A couple of years later Oracle entered with Exadata… a product competitive enough to keep the Oracle faithful on an Oracle product… but nothing really special otherwise.

Teradata eventually added a columnar feature that matched Greenplum… and Greenplum focussed away from the data warehouse space. Netezza could not match the power of columnar and could not get there so they fell away.

At this point Teradata was more-or-less back on top… although Greenplum and the other chipped away based on price. In addition, Hadoop entered the market and ate away at Teradata’s dominance in the Big Data space. The impact of Hadoop is well documented in this blog.

Three-to-four years ago SAP introduced HANA and the whole market gasped. HANA was delivering 1000X performance using columnar formats, memory to eliminate I/O, and bare-metal techniques that effectively loaded data into the processor in full cache lines.

Unfortunately, SAP did not take advantage of their significant lead in the general database markets. They focussed on their large installed base of customers… pricing HANA in a way that generated revenue but did not allow for much growth in market share. Maybe this was smart… maybe not… I was not privy to the debate.

Now Oracle has responded with in-memory columnar capability and IBM has introduced BLU. We might argue over which implementation is best… but clearly whatever lead SAP HANA held is greatly diminished. Further, HANA pricing makes it a very tough sell outside of its implementation inside the SAP Business Suite.

Teradata has provided a memory-based cache under its columnar capabilities… but this is not at the same level of sophistication as the HANA, 12c, BLU technologies which compute directly against compressed columnar data.

Hadoop is catching up slowly and we should expect that barring some giant advance from the commercial space that they will reach parity in the next 5 years or so (the will claim parity sooner… but if we require all of the capabilities offered to be present there is just no way to produce mature software any faster than 5 years).

Interestingly there is one player who seems to be advancing the state of the art. Greenplum has rolled out a best-in-class optimizer with Orca… and now has acquired Quickstep which may provide the state-of-the-art in bare metal columnar computing. When these come together Greenplum could once again bounce to the top of the performance, and the price/performance, stack. In addition, Greenplum has skinnied down and is running on an open source business model. They are very Hadoop-friendly.

It will be interesting to see if this open-source business model provides the revenue to drive advanced development… there is not really a “community” behind Greenplum development. It will also be interesting to see if the skinny business model will allow for the deployment of an enterprise-level sales force… but it just might. If Pivotal combines this new technology with a focus on the large EDW market… they may become a bigger player.

Note that was sort of dumb-luck that I posted about how Hadoop might impact revenues of big database players like Teradata right before Teradata posted a loss… but do not over think this and jump to the conclusion that Teradata is dying. They are the leader in their large space. They have great technology and they more-or-less keep up with the competition. But skinnier companies can afford to charge less and Teradata, who grew up in the days of big enterprise software, will have to skinny down like Greenplum. It will be much harder for Teradata than it was for Greenplum… and both companies will struggle with profitability for a while. But it is these technology and market dynamics that give us all something to think about, blog about, and talk about over beers…

Trends in Systems Architecture and the HANA Platform

This post is a follow on to the video on Systems Architecture here (video). I’ll describe how HANA, the Platform not the Database, is cutting fat from the landscape. If you have not seen the video recently it will be worth a quick look as I will only briefly recap it here.

The video, kindly sponsored by Intel, suggests that the distributed architecture we developed to collect lots of very small microprocessors into a single platform in order to solve big enterprise sized problems is evolving as multi-core microprocessors become ever more capable. In my opinion, we are starting to collapse the distributed architecture into something different.

DBFB SAP Distributed Landscape v1
Figure 1. A Distributed SAP Landscape

Figure 1 shows how we might deploy a series of servers to implement an SAP software landscape. Note that if we removed a couple of pieces the picture represents a platform for just about any enterprise solution. In other words, it is not just about SAP ERP solutions. I have left this in so that companies with SAP in place can see where they may be headed.

Figure 2 shows the same landscape deployed in the HANA Platform. It is a very lean deployment with all of the fat of inter-operating systems combination replaced by processes communicating over lightweight thread boundaries. Squeezing the fat out will provide significant performance benefits over the distributed deployment… it will be 2-5 times faster easily… and 2-3 times faster even if you deployed the landscape on a single server in a series of virtual machines as you have to add in the overhead of a VM plus the cost of communication over virtual networks.

DBFB SAP Collapsed Landscape v1
Figure 2. The Landscape Collapsed into threads on the HANA Platform

The HANA Platform is very powerful stuff… and if you had the choice between SAP and a competitor, the ability to match the performance on 1/2 the hardware might provide an overwhelming advantage to HANA.Note that in this evaluation I have not included any advantages from the performance kick due to the HANA Database… so the SAP story is better still.

Those of you who heard me describe HANA while I was at SAP know that this is not a new story. It was to be the opening of the shelved book on HANA. So this post satisfies my longstanding promise to get this story documented even though the book never generated much executive enthusiasm.

The trend in systems architecture outlined in the video is being capitalized by SAP as the HANA Platform… and the advantage provided is significant. You should expect SAP customers to rapidly move to HANA to take advantage of this tight, high-performance landscape.

But the software world does not stand still for long… and there are already options emerging that approximate this approach. Stay tuned…

Some HANA and Intel Videos

Here are two videos of me speaking from the 2013 Intel Developer Forum FYI.

The first has some technical detail:

The second is more of a PR pitch about Intel Hadoop:

I’m working with Intel on a new video with a pretty interesting storyline (at least I hope that you find it interesting?)… so stay tuned.

Rob

Some Database Performance Concepts

I’m working on a new idea… it may or may not pan out… but here are some concepts for your consideration… with some thoughts on their performance implications.

First a reminder… a reality check. In my experience if you POC two databases at about the same price point…and one is 30% faster than the other, 1.3X, then 50% of the time the faster DBMS will win the business. If one DBMS is 2X faster… then it will win the business 90% of the time. The 10% where the faster product loses will be because of internal politics or, for an existing application, due to the migration costs. Note that IMO it is silly to do a POC if you know up front that you will not pick the winner.

Now to the concepts… Note that these are ballpark numbers to help you think about trade-offs…

102312_2108_HopFestP1Ex1.png

The latency to start fetching data from DRAM is 100 ns… from disk it is 10M ns. If we assume that a smart RDBMS pre-fetches 80% of the data into DRAM then we can assume that an in-memory DBMS has a 200,000X performance advantage over a disk-based system.

The latency to a Flash/SSD device is 100K-200K ns. With the same 80% pre-fetch assumption an in-memory DBMS will be 20,000X faster.

Note that neither of these models include data transfer times which will favor in-memory databases even more.

If we have a hybrid system with both disk and SSD and we assume that 90% of the reads hit the SSD and that both layers in the storage hierarchy achieve 80% pre-fetch then then the in-memory system will be 38,000X faster. If fewer than 90% of the reads hit the SSD, then the latency goes up quickly.

These numbers form the basis for selecting in-memory caches like Teradata’s Intelligent Memory option as well as in-memory offerings from IBM, Microsoft, Oracle and SAP.

For typical data warehouse workloads column compression will provide around a 2.5X performance boost over row compression. This has two implications: you will get 2.5X better performance using column storage and you will get 2.5X more data into the faster levels of your storage hierarchy… more in SSD and more in-memory.

If we assume that a typical query only touches 10% of the columns in the tables addressed… then column projection provides a 9X performance boost over a row store. Exadata does not support column projection in the storage layer… and other hybrid row-or-column systems provide it only for columnar tables.

If we assume that the average latency from the processor caches is 10ns (.5ns L1, 7ns, L2, 15ns L3) and the latency to DRAM is 100ns then an in-memory system which pre-fetches data effectively into the processor caches will be 10X faster than one which goes to DRAM. If we assume that a standard RDBMS which processes uncompressed standard data types (no vector processing) gets a 20% cache hit ratio then the advantage to a cache aware RDBMS which loads full cache lines is around 8X. HANA, BLU, and the Oracle in-memory products are cache aware and get this boost with some caveats.

BLU and the Oracle in-memory option are hybrid systems that often convert data to a row form for processing (see here for some data on Oracle). If we assume that they use the full columnar in-memory vector-based structures 50% of the time then these products will see a 4X performance boost. HANA recommends that all data be stored in a columnar form so it would often see the full 8x boost.

These vector-based processes also avoid the cost of decompression… and since they process compressed vector data they can fit more information into each cache line. There is another 20%-200% (1.2X-2X) boost here but I cannot estimate it closer than that.

Finally, the vector based processes use the high performance computing instruction sets (AVX2) offered on modern CPUs… and this provides another 10X+ boost. Again, BLU and Oracle will utilize the vector form less often than HANA so they will see a boost over products like Teradata… but not see as large a boost as HANA.

There are other features at play here… some products, like HANA, shard data in-memory to get all of the cores busy on each query. I have not been able to determine if BLU or Oracle in-memory are there yet? Note that this powerful feature will allow a single query to run as fast as possible… but the benefit is mitigated when there is a workload of multiple concurrent queries (if I have 4 cores and 4 queries running concurrently, one query per core, then the 4 queries will take only a little more time than if I run the 4 queries serially with each query using all 4 cores).

It is important to note that the Oracle In-memory option does not run in the storage component of an Exadata cluster… only on the RAC layer. It is unclear how this works with the in-memory option.

The bottom line is that in-memory systems are not all alike. You can sort of add up the multipliers to get in the ballpark on how much faster Teradata will be with the Intelligent Memory option… how much faster than that a hybrid row and vector-column system like BLU or Oracle In-Memory… and how much faster a pure in-memory system might be. One thing is for sure… these in-memory options will always make the difference in a POC… in every case including them or not will blow away the 2X rule I started with… and in every case the performance benefit will outweigh the extra cost… the price/performance is very likely to be there.

I know that is skipped my usual referencing so that you can see where I pulled these numbers from… but most of this information is buried in posts here and there on my blog… and as I stated up front… these are ballpark numbers. Hopefully you can see the sense behind them… but if you think I’m off please comment back and I’ll try to adjust…

Part 8 – How Hadooped is SQL Server PDW with Polybase?

Now for SQL Server… continuing the thread on RDBMS-Hadoop integration (Part 1Part 2, Part 3, Part 4Part 5, Part 6, Part 7) I have suggested that we could evaluate integration architecture using three criteria:

  1. How parallel are the pipes to move data between the RDBMS and the parallel file system;
  2. Is there intelligence to push down predicates; and
  3. Is there more intelligence to push down joins and other relational operators?

Before we start I will suggest a fourth criteria that will be more fully explored later when we consider networks and pipes… that is: how is data sharded/hashed/distributed as it moves from the distribution scheme in HDFS to an optimal, usually hashed, scheme in the target RDBMS. Consider Greenplum as an example… they move data in parallel as quickly as possible to the GPDB and then redistribute the data across GPDB segment nodes using scatter-gather, a very efficient distribution mechanism. We will consider how PDW Poybase manages this as part of our first criteria.

Also note… since I started this series Teradata has come out with a new capability: the QueryGrid. I will add a post to consider this separately… and in this note I will assume the older Teradata capability. This is a little unfair to Teradata and I apologize for that… but otherwise this post becomes too complex. I’ll make things right for Teradata ASAP.

Now on to Microsoft…

First, Polybase has effective parallel pipes to move data from HDFS to the parallel SQL Server instances in PDW. This matches the best capability of other products like Teradata and Greenplum in this category. But where Teradata and Greenplum move data and then redistribute it, pushing the data over a network twice, Poybase has pushed the PDW hash function down to the HDFS node so that data is distributed as it is sent. This very nice feature skips one full move of the data.

Our second criteria considers how smart the connector is in pushing down filters/predicates. Polybase uses a cost-based approach to determine whether is is less expensive to push predicates down or to move all of the data up to the PDW layer. This is a best-in-class capability.

For the 3rd criteria we ask does the architecture push down advanced functions like joins and aggregates… and does the architecture minimize data pulled up to join with semi-joins? Polybase again provides strong capabilities here pushing down joins and aggregates. Polybase does not use semi-joins, so there is room to improve here… but Microsoft clearly has this capability in their roadmap.

One final note… Polybase works with PDW but not with other SQL Server products. This limitation may be relevant in many cases.

PDW + Polybase is a strong offering… matching HANA in most aspects with HANA having a slight edge in push-down with semi-joins but with SQL Server matching this with the most sophisticated parallel data distribution capability.

References

Part 6: How Hadooped is HANA?

Now for HANA plus Hadoop… to continue this thread on RDBMS-Hadoop integration (Part 1, Part 2, Part 3, Part 4, Part 5) I have suggested that we could evaluate integration architecture using three criteria:

  1. How parallel are the pipes to move data between the RDBMS and the parallel file system;
  2. Is there intelligence to push down predicates; and
  3. Is there more intelligence to push down joins and other relational operators?

As a preface I need to include a note/apology. As you will see HANA may well have the best RDBMS-Hadoop integration in the market. I try hard not to blow foam about HANA in this blog… and I hope that the objective criteria I have devised to evaluate all of the products will keep this post credible… but please look at this post harder than most and push back if you think that I overstep.

First… surprisingly, HANA’s first release has only a single pipe to the Hadoop side. This is worrisome but easily fixed. It will negatively impact performance when large tables/files have to be moved up for processing.

But HANA includes Hadoop as a full partner in a federated data architecture using the Smart Data Access (SDA) engine inside the HANA address space. As a result, HANA not only pushes predicates but it uses cost-based optimization to determine what to push down and what to pull up. HANA interrogates the Hadoop system to gather statistics and uses the HANA optimizer to develop smart execution plans with awareness of both the speed of in-memory and the limited memory resources. When data in HANA is joined with data in Hadoop SDA effectively uses semi-joins to minimize the data pulled up.

Finally, HANA can develop execution plans that executes joins in Hadoop. This includes both joins between two Hadoop tables and joins where small in-memory tables are pushed down to execute the joins in Hadoop. The current limitation is that Hadoop files must be defined as Hive tables.

Here is the HANA execution plan for TPC-H query 19. HANA has pushed down all of the steps behind the Remote Row Scan step… so in this case the entire query including a nested loop join was pushed down. In other queries HANA will push only parts of the plan to Hadoop.

TPCH Q19 Plan

So HANA possesses a very sophisticated integration with Hadoop… with capabilities that minimize the amount of data moved based on the cost of the movement. This is where all products need to go. But without parallel pipes this sophisticated capability provides only a moderate advantage (see Part 5),

Note that this is not the ultimate in integration… there is another level… but I’ll leave some ideas for extending integration even further for my final post in the series.

Next… Part 7… considering Greenplum…

HANA, BLU, Hekaton, and Oracle 12c vs. Teradata and Greenplum – November 2013

Catch Me If You Can (musical)
(Photo credit: Wikipedia)

I would like to point out a very important section in the paper on Hekaton on the Microsoft Research site here. I will quote the section in total:

2. DESIGN CONSIDERATIONS 

An analysis done early on in the project drove home the fact that a 10-100X throughput improvement cannot be achieved by optimizing existing SQL Server mechanisms. Throughput can be increased in three ways: improving scalability, improving CPI (cycles per instruction), and reducing the number of instructions executed per request. The analysis showed that, even under highly optimistic assumptions, improving scalability and CPI can produce only a 3-4X improvement. The detailed analysis is included as an appendix. 

The only real hope is to reduce the number of instructions executed but the reduction needs to be dramatic. To go 10X faster, the engine must execute 90% fewer instructions and yet still get the work done. To go 100X faster, it must execute 99% fewer instructions. This level of improvement is not feasible by optimizing existing storage and execution mechanisms. Reaching the 10-100X goal requires a much more efficient way to store and process data. 

This is important because it confirms the difference in a Level 3 and a Level 2 columnar implementation as described here. It is just not possible for a Level 2 implementation with a row-based join engine to achieve the performance of a Level 3 implementation. This will allow the Level 3 implementations: HANA, BLU, Hekaton, and Oracle 12c to distance themselves from the Level 2 products: Teradata and Greenplum; by more than 10X… and this is a very significant advantage.

Related articles

CPUs and HW for HANA, BLU, Hekaton, and Oracle 12c

CPUs from retired computers waiting for recycl...
(Photo credit: Wikipedia)

This short post is intended to provide a quick warning regarding in-memory columnar and cpu requirements… with a longer post to follow.

When a row is inserted or bulk-loaded into a DBMS, if there are no indexes, the amount of cpu required is very small. The majority of the time is spent committing a transaction is the time to write a log record to persist the data.

When the same record is reformatted into a column the amount of processing required is significantly higher. The data must be parsed into columns, the values must be compressed, dictionaries may be updated, and the breadcrumbs that let the columnar data be regenerated into rows must be laid. Further, if the columnar structure are to be optimized then the data must be ordered… with a sort or some kind of index structure. I have seen academic papers that suggest that for an insert columnar processing may be 100X more than row processing… and you can see why this could be true (I apologize for not finding the reference… I’ll dig it up… as I recall I read it in a post some time back by Daniel Abadi).

Now let’s think about this… several vendors are suggesting that you can deploy their columnar features with no changes required… no new hardware… in-place. But this does not ring true if the new columnar feature requires 100X extra CPU cycles per row… or 50X… or 10X… unless you are running your database on an empty server.

This claim is a shot at SAP who, more honestly, suggests new hardware with high-end processors for their in-memory columnar product… but methinks it is marketing, not architecture, from these other folks.

BLU Meanies: Data In-memory

Cover of "Sgt. Pepper's Lonely Hearts Clu...
Cover via Amazon

IBM is presenting a DB2 Tech Talk that compares the BLU Accelerator to HANA. There are several mistakes and some odd thinking in the pitch so let me address the issues as a way to explain some things about HANA and about BLU. This blog will consider what data needs to be in-memory.

IBM like several others, continues to repeat a talking point along the lines of: “We believe that you should not have to fit all of you active data in memory…”. Let’s think about this…

Note that in the current release HANA has a constraint that all of the data in a single column, the entire vector that represents the data in that column, must be in-memory before it can be operated on. If the table is partitioned and partition-elimination is applied then the data in the partition for the column must be loaded in-memory. This is a real constraint that will be removed in a subsequent release… but it is not a very severe constraint if you think about it.

But let’s be clear… HANA does not require all data to be in-memory… it will read data from peripheral devices in and out as required just as BLU does.

Now what does this mean? Let’s walk through some scenarios.

First, let’s imagine a customer with 10TB of user data, per the scenario IBM discusses. Let’s not get into a whose product compresses better discussion and assume that both BLU and HANA will get 4X compression… so there is 2.5TB of user data to be processed.

Now let’s imagine a system with only a very little memory available for data. In other words, let’s configure both BLU and HANA so that they are full columnar databases, but not in-memory databases. In this case BLU would operate by doing constant I/O without constraint and HANA would fail whenever it could not fit a required column in memory. Note that HANA might not fail at all… it would depend on whether there was a large single un-partitioned column that was required.

This scenario is really silly though… HANA is an in-memory database, designed to keep data in-memory from the start… so SAP would not support this imaginary configuration. The fact that you could make BLU work out of memory is not really relevant as nowhere does IBM position, or reference, BLU as a disk-based column store add-on… you would just use DB2.

Now let’s configure a system to IBM’s specification with 400GB of memory. IBM does not really say how much of this memory is available to BLU for data… but for the sake of argument let’s ignore the system requirements and assume that BLU uses one-half, 200GB, as work space to process queries so that 200GB is available to store data in-memory. As you will see it does not really matter in this argument whether I am spot on here or not. So using IBM’s recommendation there is now a 200GB cache that can be used as data is paged in and out. Anyone who has ever used a data warehouse knows that caching does not work well for BI queries as each query touches large enough volumes of the data to flush the cache… so BLU will effectively be performing I/O for most queries and is back to being an out-of-memory columnar database. Note that this flushing issue is why the in-memory capabilities from Oracle and Teradata pin certain tables into memory. In this scenario HANA will operate exactly as BLU does with the constraint that any single column that in a compressed form exceeds 200GB will not be able to be processed.

Finally let’s configure a system with 5TB of memory per SAP’s recommendation for HANA. In this case BLU and HANA both fit all of the data in-memory… with 2.5TB of compressed user data in and 2.5TB of work space… and there is no I/O. This is an in-memory DBMS.

But according to the IBM Power 770 spec (here) there is no way to get 5TB of memory on a single p770 node… so to match HANA and eliminate all I/O they would require two nodes… but BLU cannot be deployed on a cluster… so on they would have to deploy on a single node and perform I/O on 20% of the data. The latency for SSD I/O is 200Kns and for disk it is 10Mns… for DRAM it is 100ns and HANA loads full cache lines so that the average latency is under 20ns… so the penalty paid by BLU is severe and it will never keep up with HANA.

There is more bunk around recommendations for the number of cores but I can make no sense of it at all so I do not know where to begin to debunk it. SAP recommends high-end Intel servers to run HANA. In the scenario above we would recommend multiple servers… soon enough there will be Haswell servers with 6TB of DRAM and this case will run on one node.

I have stated repeatedly that anytime a vendor presents a slide comparing their product to their competitors you should immediately throw them out… it will always be twisted. Don’t trust them. And don’t trust me as I work for SAP. But hopefully you can see some logic in my case. If you need an IMDB then you need memory. If you are short of memory then the IMDB operates like a columnar RDBMS with a memory cache. If you are running a BI query workload then you need to pin data in the cache or the system will thrash. Because of this SAP recommends that you get enough memory to get all of the data in… we recommend that you operate our in-memory database product in-memory…

This really the point of the post. The Five Minute Rule informs us about what data should be in-memory (see here). An in-memory database is designed from the bottom up to manage hot data in-memory. The in-memory add-ons being offered over legacy systems are very capable and should not be ignored… and as the price of memory drops the Five Minute Rule will suggest that data in-memory will account for and ever larger percentage of your EDW. But to offer an in-memory capability and recommend that you should keep the bulk of the data on disk is silly… and to state that your product has a competitive advantage because you do not recommend that all of the data managed by your in-memory feature be kept in-memory is silliness squared.