Here is a simple test that shows the problem. My setup is:
- DSE 1.0.3 on Ubuntu 11.04, JDK 1.6.0_29 on x86_64, installed from
the DataStax debian repo (yesterday)
- Hector 1.0-1 (from maven)
Attached is a CLI file to create the keyspace and CF, and a java file to insert
data and do some queries.
This creates the following CF:
create column family IndexTest with
key_validation_class = UTF8Type
and comparator = UTF8Type
and column_metadata = [
{column_name:year, validation_class:IntegerType, index_type: KEYS},
{column_name:month, validation_class:IntegerType, index_type: KEYS},
{column_name:day, validation_class:IntegerType, index_type: KEYS},
{column_name:hour, validation_class:IntegerType, index_type: KEYS},
{column_name:minute, validation_class:IntegerType, index_type: KEYS},
{column_name:data, validation_class:UTF8Type}
];
Then inserts 5 rows per minute value, with the following values for
year/month/day/hour/minute:
Year: 2011
Month: 1, 2
Day: 1-15
Hour: 1-23
Minute: 1-59
For a total of 203,550 rows. For queries it just picks some known values for
year/month/day/hour/minute at random and looks for rows, there should be 5 rows
per combination.
Row keys are of the form YEAR-MONTH-DAY-HOUR-MINUTE-NUM (where NUM is 1-5).
Now once that data is inserted, using the CLI I can find records such as the
following:
[default@Test] get IndexTest[2011-1-8-18-30--1];
=> (column=data,
value=xvktwirapi0qs0ta29w9rchbdc2omsuv0k2chjqp9pmaodlj9ngecllaa8eq3nnx66p591b2a06mry4rpsvkd54ji5pbxikpc6mxj4czi4nuuxgoasibjd5yk65hdtqe8a0uq3yxnw81dgq6hkx8wnbs177rwo51xtkwuhwizoc0gul92pvo6tfivjgdschd9fjzfu4v1d1uxhih3argr1mp4i1h6fqybfv2utlzdzzqczq3ruu90647prrnqwdw1zqmd46ia175a929ltx2hoz8sv6rs817zm2myhp3wekfk3flnuniqgtpth7g5fns8q3oc8qde5btivt1j99gc1h2kxjbek1p448t1hs91lh9r6yrg1douj53sn7d81bnwp4nnbmz01dbr46fae1b9ter0zljet2nl1x751no6pdt64k2mdh0un01gerfihak6vn0wdvgzuv9soji3pwgnffkw2zvm5q0jlp1uf9nmy7gzswydpxwtvc35c6jw64d,
timestamp=1320769482652005)
=> (column=day, value=8, timestamp=1320769482652002)
=> (column=hour, value=18, timestamp=1320769482652003)
=> (column=minute, value=30, timestamp=1320769482652004)
=> (column=month, value=1, timestamp=1320769482652001)
=> (column=year, value=2011, timestamp=1320769482652000)
Returned 6 results.
However a CQL query to find that same record fails:
[default@Test] get IndexTest where year=2011 and month=1 and day=8 and hour=18
and minute=30;
0 Row Returned.
[default@Test] get IndexTest where year=2011 and month=1 and day=8 and hour=18;
0 Row Returned.
[default@Test] get IndexTest where year=2011 and month=1 and day=8;
0 Row Returned.
[default@Test] get IndexTest where year=2011 and month=1;
Similar results using CQLSH:
cqlsh> select * from IndexTest where year=2011 and month=1 and day=8 and
hour=18 and minute=30;
cqlsh> select * from IndexTest where year=2011 and month=1 and day=8 and
hour=18;
cqlsh> select * from IndexTest where year=2011 and month=1 and day=8;
(no results in any of those cases).
However, some data does show up through CQL (I omitted the column data for
brevity):
[default@Test] get IndexTest where year=2011 and month=2 and day=8 and hour=18
and minute=30;
-------------------
RowKey: 2011-2-8-18-30--1
-------------------
RowKey: 2011-2-8-18-30--4
-------------------
RowKey: 2011-2-8-18-30--5
-------------------
RowKey: 2011-2-8-18-30--2
-------------------
RowKey: 2011-2-8-18-30--3
5 Rows Returned.
So it seems like (in this case), month=1 is not working, but month=2 does work
(along with the other parts of the expression). I havn't tried this a bunch of
times to see if this is always the case, but it seems to be.
When running those queries using Hector, in the debugger the QueryResult's
get() method returns null (which should have rows).
Thanks,
-nate
From: Jake Luciani [mailto:[email protected]]
Sent: Tuesday, November 08, 2011 8:56 AM
To: [email protected]
Subject: Re: Secondary index issue, unable to query for records that should be
there
Hi Nate,
Could you try running it with debug enabled on the logs? it will give more
insite into what's going on.
-Jake
On Tue, Nov 8, 2011 at 3:45 PM, Nate Sammons
<[email protected]<mailto:[email protected]>> wrote:
This is against a single server, not a cluster. Replication factor for the
keyspace is set to 1, CL is the default for Hector, which I think is QUORUM.
I'm trying to get a simple test together that shows this. Does anyone know if
multiple indexes like this are efficient?
Thanks,
-nate
From: Riyad Kalla [mailto:[email protected]<mailto:[email protected]>]
Sent: Monday, November 07, 2011 4:31 PM
To: [email protected]<mailto:[email protected]>
Subject: Re: Secondary index issue, unable to query for records that should be
there
Nate, is this all against a single Cassandra server, or do you have a ring
setup? If you do have a ring setup, what is your replicationfactor set to? Also
what ConsistencyLevel are you writing with when storing the values?
-R
On Mon, Nov 7, 2011 at 2:43 PM, Nate Sammons
<[email protected]<mailto:[email protected]>> wrote:
Hello,
I'm experimenting with Cassandra (DataStax Enterprise 1.0.3), and I've got a CF
with several secondary indexes to try out some options. Right now I have the
following to create my CF using the CLI:
create column family MyTest with
key_validation_class = UTF8Type
and comparator = UTF8Type
and column_metadata = [
-- absolute timestamp for this message, also indexed
year/month/day/hour/minute
-- index these as they are low cardinality
{column_name:messageTimestamp, validation_class:LongType},
{column_name:messageYear, validation_class:IntegerType, index_type: KEYS},
{column_name:messageMonth, validation_class:IntegerType, index_type:
KEYS},
{column_name:messageDay, validation_class:IntegerType, index_type: KEYS},
{column_name:messageHour, validation_class:IntegerType, index_type: KEYS},
{column_name:messageMinute, validation_class:IntegerType, index_type:
KEYS},
... other non-indexed columns defined
];
So when I insert data, I calculate a year/month/day/hour/minute and set these
values on a Hector ColumnFamilyUpdater instance and update that way. Then
later I can query from the command line with CQL such as:
get MyTest where messageYear=2011 and messageMonth=6 and
messageDay=1 and messageHour=13 and messageMinute=44;
etc. This generally works, however at some point queries that I know should
return data no longer return any rows.
So for instance, part way through my test (inserting 250K rows), I can query
for what should be there and get data back such as the above query, but later
that same query returns 0 rows. Similarly, with fewer clauses in the
expression, like this:
get MyTest where messageYear=2011 and messageMonth=6;
Will also return 0 rows.
???????
Any idea what could be going wrong? I'm not getting any exceptions in my
client during the write, and I don't see anything in the logs (no errors
anyway).
A second question - is what I'm doing insane? I'm not sure that performance on
CQL queries with multiple indexed columns is good (does Cassandra intelligently
use all available indexes on these queries?)
Thanks,
-nate
--
http://twitter.com/tjake
SecondaryIndexTest.cli
Description: SecondaryIndexTest.cli
SecondaryIndexTest.java
Description: SecondaryIndexTest.java
