Our performance engineer, Mujtaba Chohan has agreed to put together a
benchmark for you. We only have a four node cluster of pretty average
boxes, but it should give you an idea.
No performance impact for the attrib_id not being part of the PK since
you're not filtering on them (if I understand things correctly).
A few more questions for you:
- How many rows should be use? 1B?
- How many rows would be filtered by object_id and field_type?
- Any particular key distribution or is random fine?
- What's the minimum key size we should use for object_id and
field_type? 2 bytes each?
- Any particular kind of aggregation? count(attrib1)? sum(attrib1)? A
sample query would be helpful
Since you're upgrading, use the latest on the 0.94 branch, 0.94.7.
Thanks,
James
On 04/25/2013 04:19 PM, Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN) wrote:
James: First of all, this looks quite promising.
The table schema outlined in your other message is correct except that
attrib_id will not be in the primary key. Will that be a problem with respect
to the skip-scan filter's performance? (it doesn't seem like it...)
Could you share any sort of benchmark numbers? I want to try this out right
away, but I've to wait for my cluster administrator to upgrade us from HBase
0.92 first!
----- Original Message -----
From: [email protected]
To: [email protected]
At: Apr 25 2013 18:45:14
On 04/25/2013 03:35 PM, Gary Helmling wrote:
I'm looking to write a service that runs alongside the region servers and
acts a proxy b/w my application and the region servers.
I plan to use the logic in HBase client's HConnectionManager, to segment
my request of 1M rowkeys into sub-requests per region-server. These are
sent over to the proxy which fetches the data from the region server,
aggregates locally and sends data back. Does this sound reasonable or even
a useful thing to pursue?
This is essentially what coprocessor endpoints (called through
HTable.coprocessorExec()) basically do. (One difference is that there is a
parallel request per-region, not per-region server, though that is a
potential optimization that could be made as well).
The tricky part I see for the case you describe is splitting your full set
of row keys up correctly per region. You could send the full set of row
keys to each endpoint invocation, and have the endpoint implementation
filter down to only those keys present in the current region. But that
would be a lot of overhead on the request side. You could split the row
keys into per-region sets on the client side, but I'm not sure we provide
sufficient context for the Batch.Callable instance you provide to
coprocessorExec() to determine which region it is being invoked against.
Sudarshan,
In our head branch of Phoenix (we're targeting this for a 1.2 release in
two weeks), we've implemented a skip scan filter that functions similar
to a batched get, except:
1) it's more flexible in that it can jump not only from a single key to
another single key, but also from range to range
2) it's faster, about 3-4x.
3) you can use it in combination with aggregation, since it's a filter
The scan is chunked up by region and only the keys in each region are
sent, along the lines as you and Gary have described. Then the results
are merged together by the client automatically.
How would you decompose your row key into columns? Is there a time
component? Let me walk you through an example where you might have a
LONG id value plus perhaps a timestamp (it work equally well if you only
had a single column in your PK). If you provide a bit more info on your
use case, I can tailor it more exactly.
Create a schema:
CREATE TABLE t (key BIGINT NOT NULL, ts DATE NOT NULL, data VARCHAR
CONSTRAINT pk PRIMARY KEY (key, ts));
Populate your data using our UPSERT statement.
Aggregate over a set of keys like this:
SELECT count(*) FROM t WHERE key IN (?,?,?) AND ts > ? AND ts < ?
where you bind the ? at runtime (probably building the statement
programmatically based on how many keys you're binding.
Then Phoenix would jump around the key space of your table using the
skip next hint feature provided by filters. You'd just use the regular
JDBC ResultSet to get your count back.
If you want more info and/or a benchmark of seeking over 250K keys in a
billion row table, let me know.
Thanks,
James
