Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [rdf4j-dev] UUID and registration of the StrictEvaluationStrategy
  • From: jerven Bolleman <jerven.bolleman@sib.swiss>
  • Date: Mon, 22 Nov 2021 09:14:52 +0100
  • Arc-authentication-results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=sib.swiss; dmarc=pass action=none header.from=sib.swiss; dkim=pass header.d=sib.swiss; arc=none
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=HsJmeSE/ZIQuSfV+v1tj/+u2KegmoRO3uA503HDmEpo=; b=iXoqdiAt4P38I1MGeOXhYHr7I8DZykW045Rt2e2XeL40NFAvQrDHGLzE7NnBvBYif/0q+ViHKmp+8E8HvsfQVx+5/hLPqf3Znh6FrnJaK6hzQ1Hhv9dP4XFpBqs343jp/32fgJSEpFI3cJJYN9xbiB2n/uMVXhnvrclhSnPl1decmFXa0X8w9iOFaomvNSLK7QLum5OMgXljf+lN+RZ82DwZ7UvQQhwV0W2KARjMFTIga7AMFQ8GBfOBRhH2LAvASPymve4X51FsU1pBvwQQndloEB/y7SUMZMu601fOGHF1fpHVEjoQRVLk3oCnOn0TDGYZQlCw38oF48QaW0pQXw==
  • Arc-seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=kZxqSomMW9iSPOwYnAPhHk/G4H1Ym3Md2dE2KmmVLLJGcyNs3F79t8WP7HUjim/b4QbOCFwjX3/H15Hyy6QydKYv3f4FB++X9pdLDEKX8GZ9DknAC7EUpbdKBgNJPPDqzPZ9dRwfzusaMWkuWFXvYZp8VskFRVffeBWsXjpeXYI9VOEPEuEX4A6BLniExPC1PGW4alTZIufq1JGKgZ1Rsv4e2Qcek+7WLbVyo9ar1NK2B9I3XLlGWRAvmKI1XkV8EcObfTyILbTYDXGPCz8m8UhAolUrvtm9MbgDYMTZByXNZfr2mQSIFlD5oWqcXUZgqfwpEUld/HlPFmS0fReYuA==
  • Delivered-to: rdf4j-dev@xxxxxxxxxxx
  • List-archive: <https://www.eclipse.org/mailman/private/rdf4j-dev/>
  • List-help: <mailto:rdf4j-dev-request@eclipse.org?subject=help>
  • List-subscribe: <https://www.eclipse.org/mailman/listinfo/rdf4j-dev>, <mailto:rdf4j-dev-request@eclipse.org?subject=subscribe>
  • List-unsubscribe: <https://www.eclipse.org/mailman/options/rdf4j-dev>, <mailto:rdf4j-dev-request@eclipse.org?subject=unsubscribe>
  • User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.2.0

Hi Håvard, Jeen,

I think this is to support some serialization capabilities. However,
I think that in the opensource users there are no users for serialization except for one case.

The use case for serialization that is actively used is: The OrderIterator when writing to disk it uses the serialization protocol of the Value instances. This could be replaced by the BINARY tuple query result format. Or something store specific that is even faster and more compact.

The serialization is not used in the GroupIterator where instead the MapDB logic of a POJO serializer is used.

Regards,
Jerven



On 22/11/2021 07:51, Håvard Ottestad wrote:
Hi Jeen!

I was wondering if you could explain the need to register the evaluation strategy?

I’ve been doing some profiling and it comes up as a point of contention when creating a lot of queries. Two things essentially, one is the use of UUIDs which is fairly costly due to using the platform secure random source. The second is the need to store the eval strategy in a the cache together with the synchronization inside the cache.

https://github.com/eclipse/rdf4j/blob/fbfb00d3233b2f407c837dbbeb9fa66776cb0f75/core/queryalgebra/evaluation/src/main/java/org/eclipse/rdf4j/query/algebra/evaluation/impl/StrictEvaluationStrategy.java#L229 <https://github.com/eclipse/rdf4j/blob/fbfb00d3233b2f407c837dbbeb9fa66776cb0f75/core/queryalgebra/evaluation/src/main/java/org/eclipse/rdf4j/query/algebra/evaluation/impl/StrictEvaluationStrategy.java#L229>

Essentially I’m wondering if we can just remove the use of both the UUID and the registration?

Cheers,
Håvard

_______________________________________________
rdf4j-dev mailing list
rdf4j-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/rdf4j-dev


--

	*Jerven Tjalling Bolleman*
Principal Software Developer
*SIB | Swiss Institute of Bioinformatics*
1, rue Michel Servet - CH 1211 Geneva 4 - Switzerland
t +41 22 379 58 85
Jerven.Bolleman@sib.swiss - www.sib.swiss



Back to the top