You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ofbiz.apache.org by Adrian Crum <ad...@sandglass-software.com> on 2014/09/02 12:02:15 UTC
SequenceUtil Discussion (was: (OFBIZ-2353) SequenceUtil may generate
duplicate IDs in Load Balancing mode)
Moving this discussion to the dev mailing list since the Jira issue is
closed.
Jacques,
The description of the current design can be found here:
http://ofbiz.apache.org/docs/entity.html
where it says:
"For unique primary key generation we use an original pattern we like to
call the "Ethernet Key Generation" pattern because it uses a collision
detection mechanism to insure that multiple servers can use a single
database to get banks of unique keys in a database independent way."
So, the document is making an analogy - it is saying the design is LIKE
Ethernet collision detection. The term "CSMA/CD" is an Ethernet term,
not a software design term.
Adrian Crum
Sandglass Software
www.sandglass-software.com
On 9/2/2014 10:10 AM, Jacques Le Roux (JIRA) wrote:
>
> [ https://issues.apache.org/jira/browse/OFBIZ-2353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14118052#comment-14118052 ]
>
> Jacques Le Roux commented on OFBIZ-2353:
> ----------------------------------------
>
> Thanks for your explanation Adrian, so the known pattern in code litterature is the CSMA/CD.
>
> About performance, we should remember that the size of the bank is an important parameter here.
>
> If performance is a crucial concern, people might look at using solution like CREATE SEQUENCE... Of course not in OFBiz OOTB...
>
>> SequenceUtil may generate duplicate IDs in Load Balancing mode
>> ---------------------------------------------------------------
>>
>> Key: OFBIZ-2353
>> URL: https://issues.apache.org/jira/browse/OFBIZ-2353
>> Project: OFBiz
>> Issue Type: Bug
>> Components: framework
>> Affects Versions: Release Branch 4.0, Release Branch 09.04, Trunk
>> Reporter: Philippe Mouawad
>> Assignee: Jacopo Cappellato
>> Priority: Critical
>> Fix For: Release Branch 10.04, Upcoming Branch
>>
>> Attachments: OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353-AC.patch, OFBIZ-2353-AC.patch
>>
>>
>> If Ofbiz is deploy on 2 servers in Load Balancing Mode
>> SequenceUtil will generate duplicate IDs because synchronization is done at JVM level instead of doing it in DB.
>> A good replacement implementation would be:
>> org.hibernate.id.enhanced.TableGenerator
>> But it would involve a dependency on Hibernate
>> Philippe
>> www.ubik-ingenierie.com
>
>
>
> --
> This message was sent by Atlassian JIRA
> (v6.3.4#6332)
>
Re: SequenceUtil Discussion (was: (OFBIZ-2353) SequenceUtil may generate duplicate IDs in Load Balancing mode)
Posted by Jacopo Cappellato <ja...@hotwaxmedia.com>.
as a side note, after the recent changes, the algorithm is no more based on collision detection; that description is now outdated.
Jacopo
On Sep 2, 2014, at 12:02 PM, Adrian Crum <ad...@sandglass-software.com> wrote:
> Moving this discussion to the dev mailing list since the Jira issue is closed.
>
> Jacques,
>
> The description of the current design can be found here:
>
> http://ofbiz.apache.org/docs/entity.html
>
> where it says:
>
> "For unique primary key generation we use an original pattern we like to call the "Ethernet Key Generation" pattern because it uses a collision detection mechanism to insure that multiple servers can use a single database to get banks of unique keys in a database independent way."
>
> So, the document is making an analogy - it is saying the design is LIKE Ethernet collision detection. The term "CSMA/CD" is an Ethernet term, not a software design term.
>
> Adrian Crum
> Sandglass Software
> www.sandglass-software.com
>
> On 9/2/2014 10:10 AM, Jacques Le Roux (JIRA) wrote:
>>
>> [ https://issues.apache.org/jira/browse/OFBIZ-2353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14118052#comment-14118052 ]
>>
>> Jacques Le Roux commented on OFBIZ-2353:
>> ----------------------------------------
>>
>> Thanks for your explanation Adrian, so the known pattern in code litterature is the CSMA/CD.
>>
>> About performance, we should remember that the size of the bank is an important parameter here.
>>
>> If performance is a crucial concern, people might look at using solution like CREATE SEQUENCE... Of course not in OFBiz OOTB...
>>
>>> SequenceUtil may generate duplicate IDs in Load Balancing mode
>>> ---------------------------------------------------------------
>>>
>>> Key: OFBIZ-2353
>>> URL: https://issues.apache.org/jira/browse/OFBIZ-2353
>>> Project: OFBiz
>>> Issue Type: Bug
>>> Components: framework
>>> Affects Versions: Release Branch 4.0, Release Branch 09.04, Trunk
>>> Reporter: Philippe Mouawad
>>> Assignee: Jacopo Cappellato
>>> Priority: Critical
>>> Fix For: Release Branch 10.04, Upcoming Branch
>>>
>>> Attachments: OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353-AC.patch, OFBIZ-2353-AC.patch
>>>
>>>
>>> If Ofbiz is deploy on 2 servers in Load Balancing Mode
>>> SequenceUtil will generate duplicate IDs because synchronization is done at JVM level instead of doing it in DB.
>>> A good replacement implementation would be:
>>> org.hibernate.id.enhanced.TableGenerator
>>> But it would involve a dependency on Hibernate
>>> Philippe
>>> www.ubik-ingenierie.com
>>
>>
>>
>> --
>> This message was sent by Atlassian JIRA
>> (v6.3.4#6332)
>>
Re: SequenceUtil Discussion
Posted by Jacques Le Roux <ja...@les7arts.com>.
Le 02/09/2014 12:02, Adrian Crum a écrit :
> Moving this discussion to the dev mailing list since the Jira issue is closed.
>
> Jacques,
>
> The description of the current design can be found here:
>
> http://ofbiz.apache.org/docs/entity.html
>
> where it says:
>
> "For unique primary key generation we use an original pattern we like to call the "Ethernet Key Generation" pattern because it uses a collision
> detection mechanism to insure that multiple servers can use a single database to get banks of unique keys in a database independent way."
>
> So, the document is making an analogy - it is saying the design is LIKE Ethernet collision detection. The term "CSMA/CD" is an Ethernet term, not a
> software design term.
I don't want to go on and on on this, I was just saying it's a pattern "like" CSMA/CD, to quickly grab the idea. BTW, as pointed Jacopo we should
update that document if we keep as current...
Note that the previous implementation of the pattern was faulty, hence my hasty introduction of SELECT FOR UPDATE (in the middle of a project).
Jacques
>
> Adrian Crum
> Sandglass Software
> www.sandglass-software.com
>
> On 9/2/2014 10:10 AM, Jacques Le Roux (JIRA) wrote:
>>
>> [
>> https://issues.apache.org/jira/browse/OFBIZ-2353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14118052#comment-14118052
>> ]
>>
>> Jacques Le Roux commented on OFBIZ-2353:
>> ----------------------------------------
>>
>> Thanks for your explanation Adrian, so the known pattern in code litterature is the CSMA/CD.
>>
>> About performance, we should remember that the size of the bank is an important parameter here.
>>
>> If performance is a crucial concern, people might look at using solution like CREATE SEQUENCE... Of course not in OFBiz OOTB...
>>
>>> SequenceUtil may generate duplicate IDs in Load Balancing mode
>>> ---------------------------------------------------------------
>>>
>>> Key: OFBIZ-2353
>>> URL: https://issues.apache.org/jira/browse/OFBIZ-2353
>>> Project: OFBiz
>>> Issue Type: Bug
>>> Components: framework
>>> Affects Versions: Release Branch 4.0, Release Branch 09.04, Trunk
>>> Reporter: Philippe Mouawad
>>> Assignee: Jacopo Cappellato
>>> Priority: Critical
>>> Fix For: Release Branch 10.04, Upcoming Branch
>>>
>>> Attachments: OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353 SELECT FOR UPDATE solution.patch, OFBIZ-2353-AC.patch,
>>> OFBIZ-2353-AC.patch
>>>
>>>
>>> If Ofbiz is deploy on 2 servers in Load Balancing Mode
>>> SequenceUtil will generate duplicate IDs because synchronization is done at JVM level instead of doing it in DB.
>>> A good replacement implementation would be:
>>> org.hibernate.id.enhanced.TableGenerator
>>> But it would involve a dependency on Hibernate
>>> Philippe
>>> www.ubik-ingenierie.com
>>
>>
>>
>> --
>> This message was sent by Atlassian JIRA
>> (v6.3.4#6332)
>>
>