Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>A lot would depend on the performance requirements, consistency guarantees, number of systems etc (pretty much a standard). That said I do have a recommendation if willing to have it probabilistic, and you dont need it instantly constantly upto date. HyperLogLog, as you mentioned is a great choice. </p> <p>Each server could keep one in memory that as data is inserted it updates and periodically (ie once per minute) pushes to cassandra... or some other store. Due to the nature of HyperLogLog, when you want to query it you can do a column slice of the servers hyperlogs and combine them.</p> <pre><code>ColumnFamilyHyperLogLogs: { someMetricsCardinalityRow20130828: { Server1: HyperLogLogBlob, Server2: HyperLogLogBlob, Server3: HyperLogLogBlob } } </code></pre> <p>I would recommend clearsprings library thats worked great for us:</p> <p><a href="https://github.com/clearspring/stream-lib/blob/master/src/main/java/com/clearspring/analytics/stream/cardinality/HyperLogLog.java" rel="nofollow">https://github.com/clearspring/stream-lib/blob/master/src/main/java/com/clearspring/analytics/stream/cardinality/HyperLogLog.java</a></p> <p>It has the functionality to convert to byte array which you can use for serializing and deserializing and has methods available to combine them.</p> <p>Alternatively, something that will take a lot more space is you could just have a row for each thing, like you say unique users.</p> <pre><code>ColumnFamilyName { uniqueUserOn20130828: { "user1" : null, "user2" : null, ... } } </code></pre> <p>Then you can just call a count on the row and it will give you the exact number of unique users. This is a lot more straight forward, easier to implement but it will take a <em>lot</em> more space but with the added benefit you can look at what users were actually there that day. This can probably done pretty easy with CQL3 and sets</p>
    singulars
    1. This table or related slice is empty.
    1. PODistinct Counters in Cassandra
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USChris Lohfink
    1. USChris Lohfink
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. COThe second approach gives us unique users for a day. But I want to achieve group by functionality which means I need to multiple queries to get daily distribution. Can I achieve the same in a single query ?
      singulars
      1. PO
      1. USmac
    2. COIf using thrift you can get them all in one query with a multiget_count. Not sure the equivalent in CQL. Hadoop can be used over cassandra too if want to do this in a more adhoc way.
      singulars
      1. PO
      1. USChris Lohfink
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload