Re: Starting templated cache from data streamer.

Atri, I thought a little bit more about this. Calling "getOrCreateCache" is not a good idea when streamer is called from a client node because it will lead to creating a cache delegate on this node. But with current approach this is not necessary and client node doesn't spent memory on it. Theref

[jira] [Created] (IGNITE-1068) Backpressure control is broken for ordered messages.

Vladimir Ozerov created IGNITE-1068: --- Summary: Backpressure control is broken for ordered messages. Key: IGNITE-1068 URL: https://issues.apache.org/jira/browse/IGNITE-1068 Project: Ignite I

Re: IGNITE 1055

Thank you. Please let me know when it is committed. Regards, Atri On 30 Jun 2015 23:38, "Artiom Shutak" wrote: > I think someone from commiters will look at patch (approximately tomorrow) > and merge it if it's okey. > > -- Artem -- > > On Tue, Jun 30, 2015 at 6:13 PM, Atri Sharma wrote: > >

Re: Web config download

My comments are below... On Tue, Jun 30, 2015 at 7:01 PM, Andrey Novikov wrote: > If we try to switch from > node.js: > Need implements authorization and authentication from scratch. > I was thinking a table in H2: userId | username | pswdhash > Need implement integration with less, jade. >

Re: Web config download

>> Need implement integration with less, jade. For now JADE could be used from command line and generate html pages. Later we could replace JADE with native java template engine. Web Storm/Idea could generate css from less on the fly, no integration needed. >>We use mongo specific query by json.

Re: Web config download

If we try to switch from node.js: Need implements authorization and authentication from scratch. Need implement integration with less, jade. Need rewrite rest logic. mongo: We use mongo specific query by json. Need transform this to relation model. We store user session and account in mongo using

Re: Web config download

Agree with Alexey about abstracting database layer. Everything else must work as it does now. Sergi 2015-06-30 9:20 GMT-07:00 Alexey Kuznetsov : > I think we could continue with AngularJS and Bootstrap for frontend. > > On Tue, Jun 30, 2015 at 11:05 PM, Dmitriy Setrakyan > > wrote: > > > How ab

Re: Ignite docker and cloud images

I think it's a great idea. Nick? Dmitriy > On Jun 30, 2015, at 10:24 AM, Konstantin Boudnik wrote: > > Shall the DockerFile be checked into the source code tree then? > > Cos > >> On Sun, Jun 28, 2015 at 10:58PM, Dmitriy Setrakyan wrote: >> Igniters, >> >> Nick Tikhonov has created a docke

Re: new committers

With all these transliterated names flying around here, a bit of Texas slang surely can't hurt. :) On 30.06.2015 20:33, Konstantin Boudnik wrote: > Ah, that's great! ;) > > On Tue, Jun 30, 2015 at 12:35PM, Ognen Duzlevski wrote: >> No just wanted you to know there are people on the list who under

IGNITE-788

Hi Gianfranco, I've done a review of ignite-788. Can you please look on my review (see comment at jira and new attached patch file). Let me know what you think. -- Artem --

Re: new committers

Ah, that's great! ;) On Tue, Jun 30, 2015 at 12:35PM, Ognen Duzlevski wrote: > No just wanted you to know there are people on the list who understand what > you said! ;) > > On Tue, Jun 30, 2015 at 12:40 PM, Konstantin Boudnik wrote: > > > Anything wrong with that? ;) > > > > On Tue, Jun 30, 20

Re: IGNITE 1055

I think someone from commiters will look at patch (approximately tomorrow) and merge it if it's okey. -- Artem -- On Tue, Jun 30, 2015 at 6:13 PM, Atri Sharma wrote: > Hi Artem, > > If the patch is ok, can you please merge it in sprint 7 if its not too much > of bothering? > > Thanks and Regard

Re: new committers

No just wanted you to know there are people on the list who understand what you said! ;) On Tue, Jun 30, 2015 at 12:40 PM, Konstantin Boudnik wrote: > Anything wrong with that? ;) > > On Tue, Jun 30, 2015 at 12:07PM, Ognen Duzlevski wrote: > > Yall? Texas slang on the dev list? Seriously? :) > >

Re: new committers

Anything wrong with that? ;) On Tue, Jun 30, 2015 at 12:07PM, Ognen Duzlevski wrote: > Yall? Texas slang on the dev list? Seriously? :) > > On Tue, Jun 30, 2015 at 12:24 AM, Konstantin Boudnik wrote: > > > Guys, > > > > this is great! I am very happy to have yall aboard! Welcome! > > > > Cos >

Re: Ignite docker and cloud images

Shall the DockerFile be checked into the source code tree then? Cos On Sun, Jun 28, 2015 at 10:58PM, Dmitriy Setrakyan wrote: > Igniters, > > Nick Tikhonov has created a docker container which automatically takes user > code, builds it with maven, and deploys it into an Ignite instance before >

Re: [RESULT] [VOTE] Apache Ignite 1.2.0 release (RC2)

We have a nighly build going on builds.apache.org. It should be easy to add one to do the release's convenience binaries there as well. Thoughts? Cos On Mon, Jun 29, 2015 at 07:18PM, Branko Čibej wrote: > On 26.06.2015 14:52, Yakov Zhdanov wrote: > > Brane, mvn clean package does this. The proces

[jira] [Created] (IGNITE-1067) Binary LICENSE and NOTICE fix

Anton Vinogradov created IGNITE-1067: Summary: Binary LICENSE and NOTICE fix Key: IGNITE-1067 URL: https://issues.apache.org/jira/browse/IGNITE-1067 Project: Ignite Issue Type: Task

Re: was 1.2 officially released?

Yeah, places like mvnrepository.com are not showing 1.2.0 either... On Mon, Jun 29, 2015 at 11:47 AM, Dmitriy Setrakyan wrote: > I saw the vote closed on incubator list, but the download links are not > updated yet. > > Yakov, since you are the one handling the vote, can you please push the > re

Re: new committers

Yall? Texas slang on the dev list? Seriously? :) On Tue, Jun 30, 2015 at 12:24 AM, Konstantin Boudnik wrote: > Guys, > > this is great! I am very happy to have yall aboard! Welcome! > > Cos > > On June 29, 2015 2:04:50 AM PDT, Dmitriy Setrakyan > wrote: > >Igniters, > > > >Please welcome the ne

IGNITE 1066

Folks, I have created IGNITE 1066. Please see and let me know your comments. Regards, Atri -- Regards, Atri *l'apprenant*

Re: Starting templated cache from data streamer.

I have created IGNITE -1066 for this. Please see and comment. On Tue, Jun 30, 2015 at 12:36 PM, Atri Sharma wrote: > I think it is a good idea but I wonder if it makes sense to also notify to > the user somehow that a cache was created or not by this function. > > Why I feel that is necessary i

Re: Web config download

I think we could continue with AngularJS and Bootstrap for frontend. On Tue, Jun 30, 2015 at 11:05 PM, Dmitriy Setrakyan wrote: > How about the rest of the stack? > > On Tue, Jun 30, 2015 at 9:03 AM, Alexey Kuznetsov > > wrote: > > > If we switch for jetty - we do NOT need node.js. > > If we sw

[jira] [Created] (IGNITE-1066) Add getOrCreateCache functionality in dataStreamer

Atri Sharma created IGNITE-1066: --- Summary: Add getOrCreateCache functionality in dataStreamer Key: IGNITE-1066 URL: https://issues.apache.org/jira/browse/IGNITE-1066 Project: Ignite Issue Type:

Re: strange warning message

I fixed the code to omit the warning if nothing was actually undeployed. --Yakov 2015-06-29 10:35 GMT-07:00 Dmitriy Setrakyan : > Igniters, > > I get the following warning message in the log, even though my cache only > has primitive values and no data was removed: > > --- > [10:18:57] Clear

Re: Incorrect exception is thrown in async mode when partial update occurs.

Hi, I have uploaded patch for this. I am not sure about internal flow hence might be missing some part. Please advice if same. Please see and let me know your comments and feedback. Regards, Atri On Tue, Jun 30, 2015 at 3:18 PM, Vladimir Ozerov wrote: > Atri, > > I would say that this probl

Re: Web config download

How about the rest of the stack? On Tue, Jun 30, 2015 at 9:03 AM, Alexey Kuznetsov wrote: > If we switch for jetty - we do NOT need node.js. > If we switch to H2 - we do NOT need MongoDB. > > On Tue, Jun 30, 2015 at 10:52 PM, Dmitriy Setrakyan > > wrote: > > > We are still going to need node.js

Re: Web config download

If we switch for jetty - we do NOT need node.js. If we switch to H2 - we do NOT need MongoDB. On Tue, Jun 30, 2015 at 10:52 PM, Dmitriy Setrakyan wrote: > We are still going to need node.js, right? > > On Tue, Jun 30, 2015 at 8:49 AM, Alexey Kuznetsov > > wrote: > > > Just my thoughts. > > > >

Re: Web config download

We don't have enough users to heavy load. Heavy load cannot be a problem for WebConfig. On Tue, Jun 30, 2015 at 6:49 PM, Alexey Kuznetsov wrote: > Just my thoughts. > > We could implement some abstraction layer for backend persistence. > And in case of user download we will package with H2 as de

Re: Web config download

We are still going to need node.js, right? On Tue, Jun 30, 2015 at 8:49 AM, Alexey Kuznetsov wrote: > Just my thoughts. > > We could implement some abstraction layer for backend persistence. > And in case of user download we will package with H2 as default storage. > > But in case of heavy load

Re: Web config download

Just my thoughts. We could implement some abstraction layer for backend persistence. And in case of user download we will package with H2 as default storage. But in case of heavy load any other db could be used, but will require manual installation and configuration. We are expecting million req

Re: Web config download

And what do you suggest we use as persistence on the backend? Currently we use Mongo. D. On Tue, Jun 30, 2015 at 8:18 AM, Sergi Vladykin wrote: > And there is nothing to do from scratch here, all the html and client side > js are the same, just switch backend. > > Sergi > > 2015-06-30 8:17 GMT-

Re: Web config download

And there is nothing to do from scratch here, all the html and client side js are the same, just switch backend. Sergi 2015-06-30 8:17 GMT-07:00 Sergi Vladykin : > I think at current stage if it takes two days to switch things, then it > makes sense. We will not be able to do it later. > > We wi

Re: Web config download

I think at current stage if it takes two days to switch things, then it makes sense. We will not be able to do it later. We will need to have a separate distribution for each supported platform. And test each individually. Sergi 2015-06-30 8:05 GMT-07:00 Dmitriy Setrakyan : > Guys, > > I would

Re: IGNITE 1055

Hi Artem, If the patch is ok, can you please merge it in sprint 7 if its not too much of bothering? Thanks and Regards, Atri On Mon, Jun 29, 2015 at 8:40 PM, Artiom Shutak wrote: > Hi, Atri, > > I've looked at your patch and I'm ok with it in general. I have a small > comment (see jira). > >

Re: Web config download

Guys, I would like to avoid "let's drop everything and redo from scratch" discussions. I don't think they will productively get us there. The stack was chosen because it required zero code or integration on the server side and was easiest to implement. Let's try to answer how to make it downloada

Re: Web config download

> 1. How many connections Jetty + H2 could handle? > Do you have any experience? I didn't use Jetty for production deployment, but there is no problem with high load: http://stackoverflow.com/questions/3834599/is-jetty-ever-used-for-production-deployment > 2. What is the best way to store JSON in

Re: Web config download

Guys, To be honest I don't know why we choose this stack which is completely alien to java when our main product is java based. I'm sure in the future we will need better level of integration between them not only for Ignite SQL schema import. Jetty it is a battle tested server, I think we can go

Re: Ignite does not start in the absence of the network

Dmitry, I implementd fix for multicast ip finder so now node should start even if multicast is disabled. Unfortunately we could not reproduce error you get, could you please try to run node in your environmnt using latest sp-7? Thanks On Tue, Jun 30, 2015 at 1:32 AM, Dmitriy Setrakyan wrote: >

Re: Web config download

Sergey, 1. How many connections Jetty + H2 could handle? Do you have any experience? 2. What is the best way to store JSON into db and later read JSON from db? 3. Persistence feature could not be separated, because CacheTypeMetadata is a part of CacheConfiguration. Dmitriy, What do you think a

Re: New implementations release process.

Grazie -:) Vishal Garg Sent from my iPhone > On Jun 30, 2015, at 3:13 AM, Gianfranco Murador > wrote: > > Done. I've assegned the integration with Flume to you! > Thank you too, Gianfranco > > > 2015-06-30 11:11 GMT+02:00 Vishal Garg : > >> Gianfranco, >> Great. It turns out it works simila

Re: Web config download

I doubt that people will install docker to run WebConfig. If we expect running WebConfig on user machine node.js + mongoDB is a bad choose. Most of our users are java developers, the plain way is to use Jetty + H2 and pack whole WebConfig to single jar. User will be start it using "java -jar webCo

Re: Web config download

What about docker image? We will create docker image with installed nodejs, mongo and Ignite Web Config. Will this solve this problem? On Mon, Jun 29, 2015 at 4:07 PM, Dmitriy Setrakyan wrote: > From my standpoint, I would like the web-config-download to come as a zip > file without any extra in

Re: New implementations release process.

Done. I've assegned the integration with Flume to you! Thank you too, Gianfranco 2015-06-30 11:11 GMT+02:00 Vishal Garg : > Gianfranco, > Great. It turns out it works similar to Kafka. Which I happen to work on. > > You can assign it to me. > Thanks > Vishal > > On Tue, Jun 30, 2015 at 12:56 AM

Re: Incorrect exception is thrown in async mode when partial update occurs.

Atri, I would say that this problem is specific for all calls to IgniteCacheProxy.setFuture() method. It looks like we can create an extension for IgniteFutureImpl class (say, IgniteCacheFutureImpl) which will use CU.convertToCacheException(e) to convert exceptions instead of current U.convertExce

Re: [RESULT] [VOTE] Apache Ignite 1.2.0 release (RC2)

That issue has abslutely nothing to do with my question. On 30.06.2015 11:38, Anton Vinogradov wrote: > Brane, > I'm working on this issue (IGNITE-1060) > > On Mon, Jun 29, 2015 at 8:18 PM, Branko Čibej wrote: > >> On 26.06.2015 14:52, Yakov Zhdanov wrote: >>> Brane, mvn clean package does this.

Re: [RESULT] [VOTE] Apache Ignite 1.2.0 release (RC2)

Brane, I'm working on this issue (IGNITE-1060) On Mon, Jun 29, 2015 at 8:18 PM, Branko Čibej wrote: > On 26.06.2015 14:52, Yakov Zhdanov wrote: > > Brane, mvn clean package does this. The process is covered in > DEVNOTES.txt > > > > Is that what you ask about? > > No, I'm asking where (on which

Re: Incorrect exception is thrown in async mode when partial update occurs.

Vladimir, Is this problem specific to all checks of isAsync() calls? On Mon, Jun 29, 2015 at 6:29 PM, Vladimir Ozerov wrote: > Atri, > > Currently allmost all exceptions occurring in Ignite kernal are checked. > And if we want to expose them to users, we convert them to their unchecked > counte

IGNITE-536 - Kinesis

Keith Ly, I noticed you had some questions about it after I assigned it to myself. But, if you had already started. Please let me know. Vishal

[jira] [Created] (IGNITE-1065) CacheConfiguration instance from any of the setter methods looses Generics information.

Artem Shutak created IGNITE-1065: Summary: CacheConfiguration instance from any of the setter methods looses Generics information. Key: IGNITE-1065 URL: https://issues.apache.org/jira/browse/IGNITE-1065

Re: New implementations release process.

Gianfranco, Great. It turns out it works similar to Kafka. Which I happen to work on. You can assign it to me. Thanks Vishal On Tue, Jun 30, 2015 at 12:56 AM, Gianfranco Murador < murador.gianfra...@gmail.com> wrote: > Hi Vishal, > I like to work on Apache Storm , so can we do the opposite ? Le

Re: CacheConfiguration chaining

On Tue, Jun 30, 2015 at 2:06 AM, Nikolay Tikhonov wrote: > For arguments is not required to use specific type now and if we change > setters a compatibility on compilation level with previous versions might > be broken in methods *setEvictionFilter(..), setCacheStoreFactory(..) and > setIntercept

Re: CacheConfiguration chaining

For arguments is not required to use specific type now and if we change setters a compatibility on compilation level with previous versions might be broken in methods *setEvictionFilter(..), setCacheStoreFactory(..) and setInterceptor(..)*. On Tue, Jun 30, 2015 at 12:26 AM, Dmitriy Setrakyan wrot

Re: New implementations release process.

Hi Vishal, I like to work on Apache Storm , so can we do the opposite ? Let me know if that's okay. Regards, Gianfranco 2015-06-29 23:12 GMT+02:00 Vishal Garg : > Gian, > Is it possible for you to tackle - > > Implement IgniteFlumeStreamer to steam data form Apache Flume >

Re: Ignite does not start in the absence of the network

I tried to start node with LAN and Wifi disabled and it works for me, but we can implement fix for your case: if MulticastIpFinder fails to initialize multicast socket it can just print warning and if it does not have configured addresse return single address "localhost:", in this case nodes with d

Re: Starting templated cache from data streamer.

I think it is a good idea but I wonder if it makes sense to also notify to the user somehow that a cache was created or not by this function. Why I feel that is necessary is because if we are creating caches most of the time in the function to the user opaquely, it may lead to user questioning the