Mailing-List: contact dev-help@kafka.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@kafka.apache.org
Subject: Re: [VOTE] KIP-155 Add range scan for windowed state stores
To: dev@kafka.apache.org
References: <CA+rPSbZs9yLkUsXkrR1w3_oejf4d1D6SE6hkV50s4fNZKM4s8g@mail.gmail.com>
 <CAHwHRrUbocp9KTV9xgYmZMRwm-0hC-c4f5JfSVADJDTw0ZB0cQ@mail.gmail.com>
 <CAF7WS+oOGDMapRCuSEu7DvHRABMnY67wZZbRGx+mj41DLuumuQ@mail.gmail.com>
 <CAG+EsUdvLUQRWBha49c9KBbzOJq_GUDp8hKSODakpH4jtorAPQ@mail.gmail.com>
From: Michal Borowiecki <michal.borowiecki@openbet.com>
Message-ID: <af7a5e28-8de6-9934-f575-3a85cc0f4b07@openbet.com>
Date: Wed, 10 May 2017 20:25:47 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101
 Thunderbird/45.8.0
MIME-Version: 1.0
In-Reply-To: <CAG+EsUdvLUQRWBha49c9KBbzOJq_GUDp8hKSODakpH4jtorAPQ@mail.gmail.com>
Content-Type: multipart/alternative;
 boundary="------------E1BB3F3BB22F5585E1D8DEF7"
archived-at: Wed, 10 May 2017 19:25:58 -0000

--------------E1BB3F3BB22F5585E1D8DEF7
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit

Apologies, I missed the discussion (or lack thereof) about the return 
type of:

WindowStoreIterator<KeyValue<K, V>> fetch(K from, K to, long timeFrom, 
long timeTo)


WindowStoreIterator<V> (as the KIP mentions) is a subclass of 
KeyValueIterator<Long, V>

KeyValueIterator<K,V> has the following method:

/** * Peek at the next key without advancing the iterator * @return the 
key of the next value that would be returned from the next call to next 
*/ K peekNextKey();

Given the type in this case will be Long, I assume what it would return 
is the window timestamp of the next found record?


In the case of WindowStoreIterator<V> fetch(K key, long timeFrom, long 
timeTo);
all records found by fetch have the same key, so it's harmless to return 
the timestamp of the next found window but here we have varying keys and 
varying windows, so won't it be too confusing?

KeyValueIterator<Windowed<K>, V> (as in the proposed 
ReadOnlySessionStore.fetch) just feels much more intuitive.

Apologies again for jumping onto this only once the voting has already 
begun.
Thanks,
Michał

On 10/05/17 20:08, Sriram Subramanian wrote:
> +1
>
> On Wed, May 10, 2017 at 11:42 AM, Bill Bejeck <bbejeck@gmail.com> wrote:
>
>> +1
>>
>> Thanks,
>> Bill
>>
>> On Wed, May 10, 2017 at 2:38 PM, Guozhang Wang <wangguoz@gmail.com> wrote:
>>
>>> +1. Thank you!
>>>
>>> On Wed, May 10, 2017 at 11:30 AM, Xavier Léauté <xavier@confluent.io>
>>> wrote:
>>>
>>>> Hi everyone,
>>>>
>>>> Since there aren't any objections to this addition, I would like to
>> start
>>>> the voting on KIP-155 so we can hopefully get this into 0.11.
>>>>
>>>> https://cwiki.apache.org/confluence/display/KAFKA/KIP+
>>>> 155+-+Add+range+scan+for+windowed+state+stores
>>>>
>>>> Voting will stay active for at least 72 hours.
>>>>
>>>> Thank you,
>>>> Xavier
>>>>
>>>
>>>
>>> --
>>> -- Guozhang
>>>


--------------E1BB3F3BB22F5585E1D8DEF7--