cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Carlos Alonso <i...@mrcalonso.com>
Subject Re: Strategy for dividing wide rows beyond just adding to the partition key
Date Mon, 14 Mar 2016 09:54:04 GMT
Hi.

+1 to this @Jack's sentence 'Generally, Cassandra is ideal for only two use
cases (access patterns really): 1) retrieval by a specific key, and 2)
retrieval of a relatively narrow slice of contiguous data, beginning with a
specific key.'

So I think you're modelling it properly (to have fairly narrow rows). I
think you can then store in another table the initial bucket for a sensor
and either don't have the end one (taking advantage that Cassandra is very
quick at finding empty partitions) and query until today. Or, given that
your bucketing is per week, only update the 'last partition' entry for a
sensor if we're really one week after the latest saved. That will generate
one single tombstone per sensor and that doesn't sound scary I think.

On the other hand. Did you considered offloading the historical data to a
better data warehouse?

Regards

Carlos Alonso | Software Engineer | @calonso <https://twitter.com/calonso>

On 12 March 2016 at 16:59, Jack Krupansky <jack.krupansky@gmail.com> wrote:

> Generally, secondary indexes are not recommended in Cassandra. Query
> tables and/or materialized views are the recommended alternative. But it
> all depends on the specific nature of the queries and the cardinality of
> the data.
>
> Generally, Cassandra is ideal for only two use cases (access patterns
> really): 1) retrieval by a specific key, and 2) retrieval of a relatively
> narrow slice of contiguous data, beginning with a specific key.
>
> Bulk retrieval is not a great access pattern for Cassandra. The emphasis
> is on being a database (that's why CQL is so similar to SQL) rather than a
> raw data store.
>
> Sure, technically you can do bulk retrieval, but essentially that requires
> modeling and accessing using relatively narrow slices.
>
> Closing the circle, Cassandra is always enhancing its capabilities and
> there is indeed that effort underway to support wider rows, but the
> emphasis of modeling still needs to be centered on point queries and narrow
> contiguous slices.
>
> Even with Spark and analytics that may indeed need to do a full scan of a
> large amount of data, the model needs to be that the big scan is done in
> small chunks.
>
>
> -- Jack Krupansky
>
> On Sat, Mar 12, 2016 at 10:23 AM, Jason Kania <jason.kania@ymail.com>
> wrote:
>
>> Our analytics currently pulls in all the data for a single sensor reading
>> as we use it in its entirety during signal processing. We may add secondary
>> indices to the table in the future to pull in broadly classified data, but
>> right now, our only goal is this bulk retrieval.
>>
>> ------------------------------
>> *From:* Jack Krupansky <jack.krupansky@gmail.com>
>> *To:* user@cassandra.apache.org
>> *Sent:* Friday, March 11, 2016 7:25 PM
>>
>> *Subject:* Re: Strategy for dividing wide rows beyond just adding to the
>> partition key
>>
>> Thanks, that level of query detail gives us a better picture to focus on.
>> I think through this some more over the weekend.
>>
>> Also, these queries focus on raw, bulk retrieval of sensor data readings,
>> but do you have reading-based queries, such as range of an actual sensor
>> reading?
>>
>> -- Jack Krupansky
>>
>> On Fri, Mar 11, 2016 at 7:08 PM, Jason Kania <jason.kania@ymail.com>
>> wrote:
>>
>> The 5000 readings mentioned would be against a single sensor on a single
>> sensor unit.
>>
>> The scope of the queries on this table is intended to be fairly simple.
>> Here are some example queries, without 'sharding', that we would perform on
>> this table:
>>
>> SELECT "time","readings" FROM "sensorReadings"
>> WHERE "sensorUnitId"=5123 AND "sensorId"=17 AND time<=?
>> ORDER BY time DESC LIMIT 5000
>>
>> SELECT "time","readings" FROM "sensorReadings"
>> WHERE "sensorUnitId"=5123 AND "sensorId"=17 AND time>=?
>> ORDER BY time LIMIT 5000
>>
>> SELECT "time","readings" FROM "sensorReadings"
>> WHERE "sensorUnitId"=5123 AND "sensorId"=17 AND time<=? AND
>> classification=?
>> ORDER BY time DESC LIMIT 5000
>>
>> where 'classification' is secondary index that we expect to add.
>>
>> In some cases, we have to revisit all values too so a complete table scan
>> is needed:
>>
>> SELECT "time","readings" FROM "sensorReadings"
>>
>> Getting the "next" and "previous" 5000 readings is also something we do,
>> but is manageable from our standpoint as we can look at the range-end
>> timestamps that are returned and use those in the subsequent queries.
>>
>> SELECT "time","readings" FROM "sensorReadings"
>> WHERE "sensorUnitId"=5123 AND "sensorId"=17 AND time>=? AND time<=?
>> ORDER BY time LIMIT 5000
>>
>> Splitting the bulk content out of the main table is something we
>> considered too but we didn't find any detail on whether that would solve
>> our timeout problem. If there is a reference for using this approach, it
>> would be of interest to us to avoid any assumptions on how we would
>> approach it.
>>
>> A question: Is the probability of a timeout directly linked to a longer
>> seek time in reading through a partition's contents? If that is the case,
>> splitting the partition keys into a separate table would be straightforward.
>>
>> Regards,
>>
>> Jason
>>
>> ------------------------------
>> *From:* Jack Krupansky <jack.krupansky@gmail.com>
>> *To:* user@cassandra.apache.org; Jason Kania <jason.kania@ymail.com>
>> *Sent:* Friday, March 11, 2016 6:22 PM
>>
>> *Subject:* Re: Strategy for dividing wide rows beyond just adding to the
>> partition key
>>
>> Thanks for the additional information, but there is still not enough
>> color on the queries and too much focus on a premature data model.
>>
>> Is this 5000 readings for a single sensor of a single sensor unit, or for
>> all sensors of a specified unit, or... both?
>>
>> I presume you want "next" and "previous" 5000 readings as well as first
>> and last, but... you will have to confirm that.
>>
>> One technique is to store the bulk of your raw sensor data in a separate
>> table and then simply store the PK of that data in your time series. That
>> way you can have a much wider row of time series (number of rows) without
>> hitting a bulk size issue for the partition. But... I don't want to jump to
>> solutions until we have a firmer handle on the query side of the fence.
>>
>> -- Jack Krupansky
>>
>> On Fri, Mar 11, 2016 at 5:37 PM, Jason Kania <jason.kania@ymail.com>
>> wrote:
>>
>> Jack,
>>
>> Thanks for the response.
>>
>> We are targeting our database design to 10000 sensor units and each
>> sensor unit has 32 sensors. We are seeing about 700 events per day per
>> sensor, each providing about 2K of data. Based on keeping each partition to
>> about 10 Mb (based on readings we saw on performance), we chose to break
>> our partitions on a weekly basis. This is possibly finer than we need as we
>> were seeing timeouts only once a single partition was about 150Mb in size
>>
>> When pulling in data, we will typically need to pull 1 to 4 months of
>> data for our analysis and will use only the sensorUnitId and sensorId to
>> uniquely identify the data source with the timeShard value used to break up
>> our partitions. We have handling to sequentially scan based on our
>> "timeShard" value, but don't have a good handle on the determination of the
>> "timeShard" portion of the partition key at read time. The data starts
>> coming in when a subscriber starts using our system and finishes when they
>> discontinue service or put the service on hold temporarily.
>>
>> When I talk about hotspots, it isn't the time series data that is the
>> concern, it is with respect to storing the maximum and minimum timeShard
>> values in another table for subsequent lookup or the cost of running the
>> current implementation of SELECT DISTINCT. We need to run queries such as
>> getting the first or last 5000 sensor readings when we don't know the time
>> frame at which they occurred so cannot directly supply the timeShard
>> portion of our partition key.
>>
>> I appreciate your input,
>>
>> Thanks,
>>
>> Jason
>>
>> ------------------------------
>> *From:* Jack Krupansky <jack.krupansky@gmail.com>
>> *To:* "user@cassandra.apache.org" <user@cassandra.apache.org>
>> *Sent:* Friday, March 11, 2016 4:45 PM
>>
>> *Subject:* Re: Strategy for dividing wide rows beyond just adding to the
>> partition key
>>
>> I'll stay away from advising on a specific schema per se, but I'll stick
>> to the advice that you need to make sure that your queries are depending
>> solely on the columns of the primary key or relatively short slices/scans,
>> rather than run the risk of very long scans or having to process multiple
>> partitions for a single query. That's canned to some extent, but still
>> essential.
>>
>> Of course we generally wish to avoid hotspots, but with time series they
>> are unavoidable. I mean, sure you could place successive events at separate
>> partitions, but then you can't do any kind of scanning/slicing.
>>
>> But, events for separate sensors are not true hotspots in the traditional
>> sense - unless you have only a single sensor/unit.
>>
>> After considering your queries, the next step is to consider the
>> cardinality of your data - how many sensors, how many units, rate of
>> events, etc. That will feedback into queries as well, such as how big a
>> slice or scan might be, as well as sizing of partitions.
>>
>> So, how many sensor units do you expect, how many sensors per unit, and
>> expected rate of events per sensor?
>>
>> Try not to jump too quickly to specific solutions - there really is a
>> method to understanding all of this other stuff upfront.
>>
>> -- Jack Krupansky
>>
>> On Thu, Mar 10, 2016 at 12:39 PM, Jason Kania <jason.kania@ymail.com>
>> wrote:
>>
>> Jack,
>>
>> Thanks for the response. I don't think I provided enough information and
>> used the wrong terminology as your response is more the canned advice is
>> response to Cassandra antipatterns.
>>
>> To make this clearer, this is what we are doing:
>>
>> create table sensorReadings (
>> sensorUnitId int,
>> sensorId int,
>> time timestamp,
>> timeShard int,
>> readings blob,
>> primary key((sensorUnitId, sensorId, timeShard), time);
>>
>> where timeShard is a combination of year and week of year
>>
>> For known time range based queries, this works great. However, the
>> specific problem is in knowing the maximum and minimum timeShard values
>> when we want to select the entire range of data. Our understanding is that
>> if we update another related table with the maximum and minimum timeShard
>> value for a given sensorUnitId and sensorId combination, we will create a
>> hotspot and lots of tombstones. If we SELECT DISTINCT, we get a huge list
>> of partition keys for the table because we cannot reduce the scope with a
>> where clause.
>>
>> If there is a recommended pattern that solves this, we haven't come
>> across it.
>>
>> I hope makes the problem clearer.
>>
>> Thanks,
>>
>> Jason
>>
>> ------------------------------
>> *From:* Jack Krupansky <jack.krupansky@gmail.com>
>> *To:* user@cassandra.apache.org; Jason Kania <jason.kania@ymail.com>
>> *Sent:* Thursday, March 10, 2016 10:42 AM
>> *Subject:* Re: Strategy for dividing wide rows beyond just adding to the
>> partition key
>>
>> There is an effort underway to support wider rows:
>> https://issues.apache.org/jira/browse/CASSANDRA-9754
>>
>> This won't help you now though. Even with that improvement you still may
>> need a more optimal data model since large-scale scanning/filtering is
>> always a very bad idea with Cassandra.
>>
>> The data modeling methodology for Cassandra dictates that queries drive
>> the data model and that each form of query requires a separate table
>> ("query table.") Materialized view can automate that process for a lot of
>> cases, but in any case it does sound as if some of your queries do require
>> additional tables.
>>
>> As a general proposition, Cassandra should not be used for heavy
>> filtering - query tables with the filtering criteria baked into the PK is
>> the way to go.
>>
>>
>> -- Jack Krupansky
>>
>> On Thu, Mar 10, 2016 at 8:54 AM, Jason Kania <jason.kania@ymail.com>
>> wrote:
>>
>> Hi,
>>
>> We have sensor input that creates very wide rows and operations on these
>> rows have started to timeout regulary. We have been trying to find a
>> solution to dividing wide rows but keep hitting limitations that move the
>> problem around instead of solving it.
>>
>> We have a partition key consisting of a sensorUnitId and a sensorId and
>> use a time field to access each column in the row. We tried adding a time
>> based entry, timeShardId, to the partition key that consists of the year
>> and week of year during which the reading was taken. This works for a
>> number of queries but for scanning all the readings against a particular
>> sensorUnitId and sensorId combination, we seem to be stuck.
>>
>> We won't know the range of valid values of the timeShardId for a given
>> sensorUnitId and sensorId combination so would have to write to an
>> additional table to track the valid timeShardId. We suspect this would
>> create tombstone accumulation problems given the number of updates required
>> to the same row so haven't tried this option.
>>
>> Alternatively, we hit a different bottleneck in the form of SELECT
>> DISTINCT in trying to directly access the partition keys. Since SELECT
>> DISTINCT does not allow for a where clause to filter on the partition key
>> values, we have to filter several hundred thousand partition keys just to
>> find those related to the relevant sensorUnitId and sensorId. This problem
>> will only grow worse for us.
>>
>> Are there any other approaches that can be suggested? We have been
>> looking around, but haven't found any references beyond the initial
>> suggestion to add some sort of shard id to the partition key to handle wide
>> rows.
>>
>> Thanks,
>>
>> Jason
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>

Mime
View raw message