You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@chemistry.apache.org by "krzysztoffzielinski@gmail.com" <kr...@gmail.com> on 2017/11/02 18:25:31 UTC
Performance problems
Hello All
Do you experience any performance problems while talking to DMS repo using open cmis library?
For example, when you make a call session.queryObjects() it fetches the data making additional calls to repository (discoveryService.query()). Any time each of the calls is made, we need to authenticate plus have the request processed by document repository. This increase the overall time to complete a task.
Is there a way to optimize these? like fetch data as a whole in one call, rather than splitting it to separate calls?
Do I need to configure cmis session to make efficient calls?
Kind Regards
Krzysztof
Re: Performance problems
Posted by "krzysztoffzielinski@gmail.com" <kr...@gmail.com>.
Yeah. No miracles happen.
On 2017-11-03 11:13, Florian Müller <fm...@apache.org> wrote:
> The exception is returned by the server. If the server doesn't support
> it, the client has to work around it.
> Sorry, there is no magic on the client side.
>
> - Florian
>
>
> > thanks, Florian. I understand better now how query works.
> >
> > The problem is for each returned result we obtain all versions
> > seperately, which impacts performance. Seems like we cannot go
> > differently about this, as when we set searchAllVersions to true we
> > get an exception:
> >
> > CmisInvalidArgumentException: Search all version is not supported!
> >
> > Is there other way to query results with all versions, rather than
> > lazy load them?
> >
> > Krzysztof
> >
> > On 2017-11-02 19:11, Florian Müller <fm...@apache.org> wrote:
> >> Hi Krzysztof,
> >>
> >> see this page:
> >> https://chemistry.apache.org/docs/cmis-samples/samples/operation-context/index.html#list-batch-size
> >>
> >> - Florian
> >>
> >>
> >> > Hello All
> >> >
> >> > Do you experience any performance problems while talking to DMS repo using open cmis library?
> >> >
> >> > For example, when you make a call session.queryObjects() it fetches the data making additional calls to repository (discoveryService.query()). Any time each of the calls is made, we need to authenticate plus have the request processed by document repository. This increase the overall time to complete a task.
> >> >
> >> > Is there a way to optimize these? like fetch data as a whole in one call, rather than splitting it to separate calls?
> >> >
> >> > Do I need to configure cmis session to make efficient calls?
> >> >
> >> > Kind Regards
> >> > Krzysztof
> >> >
> >>
>
Re: Performance problems
Posted by Florian Müller <fm...@apache.org>.
The exception is returned by the server. If the server doesn't support
it, the client has to work around it.
Sorry, there is no magic on the client side.
- Florian
> thanks, Florian. I understand better now how query works.
>
> The problem is for each returned result we obtain all versions
> seperately, which impacts performance. Seems like we cannot go
> differently about this, as when we set searchAllVersions to true we
> get an exception:
>
> CmisInvalidArgumentException: Search all version is not supported!
>
> Is there other way to query results with all versions, rather than
> lazy load them?
>
> Krzysztof
>
> On 2017-11-02 19:11, Florian Müller <fm...@apache.org> wrote:
>> Hi Krzysztof,
>>
>> see this page:
>> https://chemistry.apache.org/docs/cmis-samples/samples/operation-context/index.html#list-batch-size
>>
>> - Florian
>>
>>
>> > Hello All
>> >
>> > Do you experience any performance problems while talking to DMS repo using open cmis library?
>> >
>> > For example, when you make a call session.queryObjects() it fetches the data making additional calls to repository (discoveryService.query()). Any time each of the calls is made, we need to authenticate plus have the request processed by document repository. This increase the overall time to complete a task.
>> >
>> > Is there a way to optimize these? like fetch data as a whole in one call, rather than splitting it to separate calls?
>> >
>> > Do I need to configure cmis session to make efficient calls?
>> >
>> > Kind Regards
>> > Krzysztof
>> >
>>
Re: Performance problems
Posted by "krzysztoffzielinski@gmail.com" <kr...@gmail.com>.
thanks, Florian. I understand better now how query works.
The problem is for each returned result we obtain all versions seperately, which impacts performance. Seems like we cannot go differently about this, as when we set searchAllVersions to true we get an exception:
CmisInvalidArgumentException: Search all version is not supported!
Is there other way to query results with all versions, rather than lazy load them?
Krzysztof
On 2017-11-02 19:11, Florian Müller <fm...@apache.org> wrote:
> Hi Krzysztof,
>
> see this page:
> https://chemistry.apache.org/docs/cmis-samples/samples/operation-context/index.html#list-batch-size
>
> - Florian
>
>
> > Hello All
> >
> > Do you experience any performance problems while talking to DMS repo using open cmis library?
> >
> > For example, when you make a call session.queryObjects() it fetches the data making additional calls to repository (discoveryService.query()). Any time each of the calls is made, we need to authenticate plus have the request processed by document repository. This increase the overall time to complete a task.
> >
> > Is there a way to optimize these? like fetch data as a whole in one call, rather than splitting it to separate calls?
> >
> > Do I need to configure cmis session to make efficient calls?
> >
> > Kind Regards
> > Krzysztof
> >
>
Re: Performance problems
Posted by Florian Müller <fm...@apache.org>.
Hi Krzysztof,
see this page:
https://chemistry.apache.org/docs/cmis-samples/samples/operation-context/index.html#list-batch-size
- Florian
> Hello All
>
> Do you experience any performance problems while talking to DMS repo using open cmis library?
>
> For example, when you make a call session.queryObjects() it fetches the data making additional calls to repository (discoveryService.query()). Any time each of the calls is made, we need to authenticate plus have the request processed by document repository. This increase the overall time to complete a task.
>
> Is there a way to optimize these? like fetch data as a whole in one call, rather than splitting it to separate calls?
>
> Do I need to configure cmis session to make efficient calls?
>
> Kind Regards
> Krzysztof
>