Yes, I say that.
To my understanding you wanted to avoid scan, so i thought you might
denormalize a bit more with another table. If you want to keep just one
of the two, your usage pattern will help you with your choice.
If you'll stick for a scan on your actual table, you'll have a big
number of columns so a bloom filter will help out for sure.
AFAIK second indexes can't be put to columns whose name isn't fixed. for
instance you might have a table with productids as keys and a column
name "cluster" whose value is the clusterid, and then second-index on
it. Then you can scan keys on the value of that well-known column.
But I might be totally wrong about this last one. Hope devs will comment
on this one.
On 12/10/10 1:08 PM, Gökhan Çapan wrote:
> Claudio,
> You say that I can flip the data, right?
> If I understand your suggestion correctly, then getting products that a
> cluster includes will be the problem.
>
> Actually, in google's BigTable paper, their example table is similar, also
> their example code to retrieve data:
> Scanner scanner(T);
> ScanStream *stream;
> stream = scanner.FetchColumnFamily("anchor");
> stream->SetReturnAllVersions();
> scanner.Lookup("com.cnn.www");
> for (; !stream->Done(); stream->Next()) {
> printf("%s %s %lld %s\n",
> scanner.RowName(),
> stream->ColumnName(),
> stream->MicroTimestamp(),
> stream->Value());
> }
>
> I was just thinking of making this scan faster within this table, maybe with
> a BloomFilter.
>
> Another table with productId - clusters row will be the next option.
>
> Thank you.
>
>
> On Fri, Dec 10, 2010 at 1:52 PM, Claudio Martella <
> [email protected]> wrote:
>
>> What about a thin table? rowkey:productid columname:clusterid?
>>
>>
>> On 12/10/10 10:52 AM, Gökhan Çapan wrote:
>>> Hi,
>>>
>>> We have the output of a clustering algorithm in an hbase table which has
>> the
>>> following structure:
>>>
>>> {NAME => 'clusters', FAMILIES => [{NAME => 'products', COMPRESS
>>> true
>>> ION => 'NONE', VERSIONS => '3', TTL => '2147483647', BLOCKSIZE =>
>>> '655
>>> 36', IN_MEMORY => 'false', BLOCKCACHE => 'true'}]}
>>>
>>> row ids are cluster ids.
>>> Columns in products column family are the id of the products.
>>>
>>> an example row is:
>>> 1-1000936175-1879240683-185 column=products:21840054,
>>> timestamp=1291817353183, value=\x00\x00\x00\x01
>>>
>>> 1-1000936175-1879240683-185 column=products:23194179,
>>> timestamp=1291817353183, value=\x00\x00\x00\x01
>>>
>>> 1-1000936175-1879240683-185 column=products:23585765,
>>> timestamp=1291817353183, value=\x00\x00\x00\x01
>>>
>>> 1-1000936175-1879240683-185 column=products:24544087,
>>> timestamp=1291817353183, value=\x00\x00\x00\x01
>>>
>>>
>>>
>>> When we want to determine which clusters a product belongs to, we
>> perform a
>>> scan over the table using column,
>>>
>>> e.g.
>>>
>>> Scan s = new Scan();
>>> s.addColumn(Bytes.toBytes("products"), Bytes.toBytes("24659517"));
>>> ResultScanner scanner = table.getScanner(s);
>>>
>>> I am not sure this is the best way, it is slow, could you suggest a
>> faster
>>> way to determine such rows?
>>> Is there a secondary index implementation that we can add to a column
>> family
>>> after adding data to table?
>>>
>>
>> --
>> Claudio Martella
>> Digital Technologies
>> Unit Research & Development - Analyst
>>
>> TIS innovation park
>> Via Siemens 19 | Siemensstr. 19
>> 39100 Bolzano | 39100 Bozen
>> Tel. +39 0471 068 123
>> Fax +39 0471 068 129
>> [email protected] http://www.tis.bz.it
>>
>> Short information regarding use of personal data. According to Section 13
>> of Italian Legislative Decree no. 196 of 30 June 2003, we inform you that we
>> process your personal data in order to fulfil contractual and fiscal
>> obligations and also to send you information regarding our services and
>> events. Your personal data are processed with and without electronic means
>> and by respecting data subjects' rights, fundamental freedoms and dignity,
>> particularly with regard to confidentiality, personal identity and the right
>> to personal data protection. At any time and without formalities you can
>> write an e-mail to [email protected] in order to object the processing of
>> your personal data for the purpose of sending advertising materials and also
>> to exercise the right to access personal data and other rights referred to
>> in Section 7 of Decree 196/2003. The data controller is TIS Techno
>> Innovation Alto Adige, Siemens Street n. 19, Bolzano. You can find the
>> complete information on the web site www.tis.bz.it.
>>
>>
>>
>
--
Claudio Martella
Digital Technologies
Unit Research & Development - Analyst
TIS innovation park
Via Siemens 19 | Siemensstr. 19
39100 Bolzano | 39100 Bozen
Tel. +39 0471 068 123
Fax +39 0471 068 129
[email protected] http://www.tis.bz.it
Short information regarding use of personal data. According to Section 13 of
Italian Legislative Decree no. 196 of 30 June 2003, we inform you that we
process your personal data in order to fulfil contractual and fiscal
obligations and also to send you information regarding our services and events.
Your personal data are processed with and without electronic means and by
respecting data subjects' rights, fundamental freedoms and dignity,
particularly with regard to confidentiality, personal identity and the right to
personal data protection. At any time and without formalities you can write an
e-mail to [email protected] in order to object the processing of your personal
data for the purpose of sending advertising materials and also to exercise the
right to access personal data and other rights referred to in Section 7 of
Decree 196/2003. The data controller is TIS Techno Innovation Alto Adige,
Siemens Street n. 19, Bolzano. You can find the complete information on the web
site www.tis.bz.it.
