[GitHub] cyejing opened a new issue #235: mq??????RebalanceService????
cyejing opened a new issue #235: mq??RebalanceService URL: https://github.com/apache/rocketmq/issues/235 https://github.com/apache/rocketmq/blob/a5d72be665f80487fd81e60604fac5c8e6a971ea/remoting/src/main/java/org/apache/rocketmq/remoting/common/RemotingHelper.java#L56 : http://xxx:8080/xxx ?? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
Re: GSOC - ROCKETMQ-380
HI,
Regarding this RocketMQ-380,
I have issues in running the referenced console when ever i try to enter my
personal details page is automatically redirected to home page.
Hence i'm not being able to activate the messaging service which i believe
should work.
https://mns.console.aliyun.com/?spm=5176.2020520130.0.0.4db63db5cHTXtU#/list/cn-hangzhou
>From this link when i click activate button it is redirecting to the home
page. but according to the documentation it should take me to a form to
fill my personal details. I created an account and i can login to my
account as well. Please help on this.
The other thing is what are the expected features in the new console?
Please give a short description.
Thanks in advance.
On Wed, Mar 7, 2018 at 12:40 AM, Ahmed Ifhaam
wrote:
> Ok sorry, changing on bin/runbrocker.sh worked Thank you
>
> On Wed, Mar 7, 2018 at 12:27 AM, Sohaib Iftikhar
> wrote:
>
>> Hi Ahmed,
>>
>> That file is for the nameserver. You also need to change the JAVA_OPTS for
>> the broker. Look at the fine `bin/runbroker.sh`.
>>
>> Thanks,
>> Sohaib
>>
>> On Tue, Mar 6, 2018 at 7:41 PM, Ahmed Ifhaam
>> wrote:
>>
>> > I tried the way you asked me to do, actually values were 4g only even
>> > without changing but ichanged it back to 2gb also and tried no effect
>> its
>> > taking 8 gb.
>> > I changed it in the file
>> > "distribution/target/apache-rocketmq/bin/runserver.sh" is it the right
>> > file
>> > to change ?
>> >
>> > Thank you
>> >
>> > On Tue, Mar 6, 2018 at 12:00 PM, Ahmed Ifhaam
>> > wrote:
>> >
>> > > THANK YOU
>> > >
>> > > On Tue, Mar 6, 2018 at 11:22 AM, yukon wrote:
>> > >
>> > >> Hi,
>> > >>
>> > >> Change this line:
>> > >>
>> > >> ```
>> > >> JAVA_OPT="${JAVA_OPT} -server -Xms8g -Xmx8g -Xmn4g"
>> > >> ```
>> > >>
>> > >> to:
>> > >>
>> > >> ```
>> > >> JAVA_OPT="${JAVA_OPT} -server -Xms4g -Xmx4g -Xmn2g"
>> > >> ```
>> > >>
>> > >> It's a Java application, broker will only generate a 4G heap now~
>> > >>
>> > >> Regards,
>> > >> yukon
>> > >>
>> > >> On Tue, Mar 6, 2018 at 1:14 PM, Ahmed Ifhaam > >
>> > >> wrote:
>> > >>
>> > >> > hi yukon,
>> > >> > As you said its taking 8gb almost thats why my system is getting
>> stuck
>> > >> > mostly. I read the scripts in the bin folder i didn't find a place
>> to
>> > >> > adjust the memory needed. I have only 8GB of ram and my hard disk
>> is
>> > >> not an
>> > >> > ssd so i need to work without swap as much as possible can you
>> help me
>> > >> on
>> > >> > this please .
>> > >> >
>> > >> > Thank you
>> > >> >
>> > >> > On Mon, Mar 5, 2018 at 5:02 PM, yukon wrote:
>> > >> >
>> > >> > > Hi,
>> > >> > >
>> > >> > > The start script will consume 8GB mem, you could adjust the
>> script
>> > >> before
>> > >> > > running broker[1].
>> > >> > >
>> > >> > > Also please try the current rocketmq console for reference[2].
>> > >> > >
>> > >> > > Let us know if you have any question.
>> > >> > >
>> > >> > > [1].
>> > >> > > https://github.com/apache/rocketmq/blob/master/
>> > >> > > distribution/bin/runbroker.sh
>> > >> > > [2].
>> > >> > > https://github.com/apache/rocketmq-externals/tree/
>> > >> > master/rocketmq-console
>> > >> > >
>> > >> > > Regards,
>> > >> > > yukon
>> > >> > >
>> > >> > > On Mon, Mar 5, 2018 at 5:33 PM, Ahmed Ifhaam <
>> > ahmedifha...@gmail.com>
>> > >> > > wrote:
>> > >> > >
>> > >> > > > Hi all,
>> > >> > > >
>> > >> > > > I'm Ahmed Ifhaam from Sri Lanka. I'm interested in
>> participating
>> > >> GSOC
>> > >> > > 2018.
>> > >> > > > For my project i'm willing to take ROCKETMQ-380
>> > >> > > > issue.
>> Since
>> > >> i'm
>> > >> > > new
>> > >> > > > to the product. Where can i start ? Please help me on this. I
>> > tried
>> > >> > > running
>> > >> > > > Rocket-MQ but it is taking a lot of memory which results my
>> system
>> > >> to
>> > >> > > hang.
>> > >> > > > I'm looking for the proper way to start solving this issue.
>> > >> > > >
>> > >> > > > Thank you.
>> > >> > > >
>> > >> > > > --
>> > >> > > > *Ahmed Ifhaam *
>> > >> > > > University of Kelaniya | MIT '19
>> > >> > > > Tel :- 94662244937 | Mob :- 94776699609
>> > >> > > > ahmedifha...@gmail.com
>> > >> > > >
>> > >> > >
>> > >> >
>> > >> >
>> > >> >
>> > >> > --
>> > >> > *Ahmed Ifhaam *
>> > >> > University of Kelaniya | MIT '19
>> > >> > Tel :- 0662244937 | Mob :- 0776699609
>> > >> > ahmedifha...@gmail.com
>> > >> >
>> > >>
>> > >
>> > >
>> > >
>> > > --
>> > > *Ahmed Ifhaam *
>> > > University of Kelaniya | MIT '19
>> > > Tel :- 0662244937 | Mob :- 0776699609
>> > > ahmedifha...@gmail.com
>> > >
>> >
>> >
>> >
>> > --
>> > *Ahmed Ifhaam *
>> > University of Kelaniya | MIT '19
>> > Tel :- 0662244937 | Mob :- 0776699609
>> > ahmedifha...@gmail.com
>> >
>>
>
>
>
> --
> *Ahmed Ifhaam *
> University of Kelaniya | MIT '19
> Tel :- 0662244937 | Mob :- 0776699609
> ahmedifha...@gmail.com
>
--
*Ahmed Ifhaam *
University of Kelaniya | MIT '19
Tel :- 0662244937 | Mob :- 0776699609
ahmedifha..
[GitHub] devsong commented on issue #235: mq??????RebalanceService????
devsong commented on issue #235: mq??RebalanceService URL: https://github.com/apache/rocketmq/issues/235#issuecomment-371693361 tcp socket address,like ip:port(eg:127.0.0.1:8080),not http url pattern This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lindzh commented on issue #235: mq??????RebalanceService????
lindzh commented on issue #235: mq??RebalanceService
URL: https://github.com/apache/rocketmq/issues/235#issuecomment-371694317
If you want to use http for nameserver discovery,please see here.
```
public static final String WS_DOMAIN_NAME =
System.getProperty("rocketmq.namesrv.domain", DEFAULT_NAMESRV_ADDR_LOOKUP);
public static final String WS_DOMAIN_SUBGROUP =
System.getProperty("rocketmq.namesrv.domain.subgroup", "nsaddr");
// http://jmenv.tbsite.net:8080/rocketmq/nsaddr
public static final String WS_ADDR = "http://"; + WS_DOMAIN_NAME +
":8080/rocketmq/" + WS_DOMAIN_SUBGROUP;
```
other wise please go to http://rocketmq.apache.org/docs/simple-example/ for
message pub/sub usage documentation.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] uptop commented on issue #234: ?tag????
uptop commented on issue #234: ?tag URL: https://github.com/apache/rocketmq/issues/234#issuecomment-371700464 ??@zhouxinyu ?message??tag???message???tag??broker filter???client filter?? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] cnhans opened a new issue #236: ???netty??????rocketmq??????netty-all?????netty-common?
cnhans opened a new issue #236: ???netty??rocketmq??netty-all?netty-common? URL: https://github.com/apache/rocketmq/issues/236 ??? 1netty-all???netty-all?jar?? 2es??es???netty?netty-common?netty?jar???jarnetty-allrocketmq??netty-commonjava.lang.NoClassDefFoundError: io/netty/util/concurrent/DefaultPromise$1 3??rocketmq?common??client?all? 4??mq? ??? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
Rocketmq support
Hi, I am from Singapore DBS bank, we are using rocketmq in our application so we are finding support team for rocketmq, please reply me if I send to right place, thanks. Thanks & Regards Xin Xin Ext: +65 6878 8531 CONFIDENTIAL NOTE: The information contained in this email is intended only for the use of the individual or entity named above and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this message in error, please immediately notify the sender and delete the mail. Thank you.
[GitHub] zhouxinyu commented on issue #234: ?tag????
zhouxinyu commented on issue #234: ?tag URL: https://github.com/apache/rocketmq/issues/234#issuecomment-371713814 Of course, any contribution is welcome. As a new feature, we recommend discussing it in dev list first. And please check out this feature [filter-by-sql92](http://rocketmq.apache.org/docs/filter-by-sql92-example/) whether meets your demand. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhouxinyu commented on issue #235: mq??????RebalanceService????
zhouxinyu commented on issue #235: mq??RebalanceService URL: https://github.com/apache/rocketmq/issues/235#issuecomment-371714190 As @devsong said, this `addr` is a TCP socket address. And the Apache RocketMQ community is global, we recommend using English as our preferred language~ This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhouxinyu closed issue #235: mq??????RebalanceService????
zhouxinyu closed issue #235: mq??RebalanceService URL: https://github.com/apache/rocketmq/issues/235 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhouxinyu commented on issue #236: ???netty??????rocketmq?????netty-all?????netty-common?
zhouxinyu commented on issue #236: ???netty??rocketmq?netty-all?netty-common? URL: https://github.com/apache/rocketmq/issues/236#issuecomment-371714537 Please **respect** our [issue template](https://github.com/apache/rocketmq/blob/master/.github/ISSUE_TEMPLATE.md) and remember the issue tracker is **ONLY** used for bug report and feature request. Any question or RocketMQ proposal please use our [mailing lists](http://rocketmq.apache.org/about/contact/). This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhouxinyu closed issue #236: ???netty??????rocketmq?????netty-all?????netty-common?
zhouxinyu closed issue #236: ???netty??rocketmq?netty-all?netty-common? URL: https://github.com/apache/rocketmq/issues/236 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
Re: 答复: [GSOC|ROCKETMQ-124] Support non-redundant message delivery mechanism
``` Personally, I find RAFT to be much simpler to implement. However, I do not expect to reinvent the wheel here. ``` That's absolutely right, no need to reinvent the wheel, there are many existing implementations for raft: https://raft.github.io/ ``` I don't think using key store to persist all the messages is a good idea. ``` Yes, store an ID is enough. On Thu, Mar 8, 2018 at 3:32 PM, Sohaib Iftikhar wrote: > Hi Dexin, > > Thank you for your suggestions. I will try to answer as much as I can and > leave the rest to the RocketMQ team. > > 1. The idea with incremental Ids is actually quite good. But @Yukon > mentioned that duplication can also be controlled by an application > (special KV Property) in which case different producers may produce the > same message that needs to deduplicated on the broker. > SessionId+IncrementalId won't work in this scenario I believe. But we can > actually switch to more efficient storage using the idea you described when > the user is not specifying these special keys. > Also I proposed storing of keys for only a fixed time interval. For all > practical purposes this would still remain constant time. [Log base 2 of > 10^10 is still just 33 :) ]. It does add the extra cost of communication > but this would be the case in both scenarios. > 2. As for consensus, the ideas I presented were pretty abstract so I > mentioned a couple of algorithms that could potentially be used. > Personally, I find RAFT to be much simpler to implement. However, I do not > expect to reinvent the wheel here. I strongly believe that in this case, we > can build upon some tested existing solution. > > > Regards, > Sohaib > > On Thu, Mar 8, 2018 at 1:31 AM, 李 德鑫 wrote: > > > Hi Sohaib, > > > > > > I‘m a student applying for GSOC too. And I've read all of your discussion > > in the mail list. > > > > I have some questions about your design, and some of the questions may > > need to be answered by RocketMQ team. So I send them here to be > discussed. > > > > I don't think using key store to persist all the messages is a good idea. > > Since MQ is based on O(1) data structure. The key store would harm the > > performance. > > > > I think we can learn from TCP protocol. > > > > In Producer-Broker Communication, we can give an incremental id for every > > message sent in the same session. And the session id should be persistent > > on the disk for producer. So the broker only need to maintain a map > between > > session id to expected message id(And this is how Kafka do it). Since > > messages are much more than producers. However, there's still a K/V store > > needed. So we have to ask RocketMQ team about how many producers in the > > same time while in practical situation. > > > > Also, the same idea in Consumer-Broker Communication. > > > > > > About consensus algorithm, I think RocketMQ should already have an > > implementation there. I don't know what it is, but maybe you can reuse > > that. Or what if you have to implement one, in my opinion, there's no > need > > to implement both Paxos and Raft. Since they solve the same kind of > > problems. > > > > > > > > Regards, > > > > Dexin > > > > > > > > 发件人: Sohaib Iftikhar > > 发送时间: 2018年3月7日 18:15:51 > > 收件人: dev@rocketmq.apache.org > > 主题: Re: [GSOC|ROCKETMQ-124] Support non-redundant message delivery > > mechanism > > > > Hi Yukon, > > > > Thanks for your reply. Yes, it would be nice to concretely define the > scope > > of this project as the doc is a bit ambitious for just a summer. Should > you > > (or anyone else) have questions/suggestions/clarifications I'd be glad > to > > discuss more details. > > > > Thanks, > > Sohaib > > > > On Wed, Mar 7, 2018 at 8:58 AM, yukon wrote: > > > > > Hi, > > > > > > Google doc is better for discussion, your design is great, now we could > > > discuss more details base on it. > > > > > > Any advice is welcome from RocketMQ community. > > > > > > Appreciate your efforts. > > > > > > Regards, > > > yukon > > > > > > On Wed, Mar 7, 2018 at 5:15 AM, Sohaib Iftikhar > > > wrote: > > > > > > > Hi, > > > > > > > > @Yukon Thank you for your reply. This clears some doubts. > > > > > > > > Sorry for the delay as I was somewhat occupied with another project. > I > > > have > > > > created an initial design doc. Email is a bit cumbersome for > feedback I > > > > wrote this document in two formats: > > > > > > > > 1. In the form of a Google document: > > > > https://docs.google.com/document/d/1KSpXGNDH0HF5E27lfKJxJnjIjPtlP > > > > 1Q-M6rj3yZde24. > > > > The document is open for comments to all users without signing in. I > > > would > > > > appreciate it if you put your name before the comment so I can > identify > > > who > > > > to follow up the discussion with. > > > > > > > > 2. As a markdown on github: > > > > https://github.com/sohaibiftikhar/rocketmq/blob/ > > > gsoc_design/gsoc_design.md > > > > . > > > > The comments for this can be made on the commit: > > > > https://githu
Re: 答复: [GSOC|ROCKETMQ-124] Support non-redundant message delivery mechanism
Hi Yukon, What do you suggest for the key store itself? Do you propose writing this ourselves or using some existing solution and writing a layer on top? Thanks, Sohaib On Fri, Mar 9, 2018 at 6:20 AM, yukon wrote: > ``` > Personally, I find RAFT to be much simpler to implement. However, I do not > expect to reinvent the wheel here. > ``` > > That's absolutely right, no need to reinvent the wheel, there are many > existing implementations for raft: https://raft.github.io/ > > ``` > I don't think using key store to persist all the messages is a good idea. > ``` > > Yes, store an ID is enough. > > > On Thu, Mar 8, 2018 at 3:32 PM, Sohaib Iftikhar > wrote: > > > Hi Dexin, > > > > Thank you for your suggestions. I will try to answer as much as I can and > > leave the rest to the RocketMQ team. > > > > 1. The idea with incremental Ids is actually quite good. But @Yukon > > mentioned that duplication can also be controlled by an application > > (special KV Property) in which case different producers may produce the > > same message that needs to deduplicated on the broker. > > SessionId+IncrementalId won't work in this scenario I believe. But we can > > actually switch to more efficient storage using the idea you described > when > > the user is not specifying these special keys. > > Also I proposed storing of keys for only a fixed time interval. For all > > practical purposes this would still remain constant time. [Log base 2 of > > 10^10 is still just 33 :) ]. It does add the extra cost of communication > > but this would be the case in both scenarios. > > 2. As for consensus, the ideas I presented were pretty abstract so I > > mentioned a couple of algorithms that could potentially be used. > > Personally, I find RAFT to be much simpler to implement. However, I do > not > > expect to reinvent the wheel here. I strongly believe that in this case, > we > > can build upon some tested existing solution. > > > > > > Regards, > > Sohaib > > > > On Thu, Mar 8, 2018 at 1:31 AM, 李 德鑫 wrote: > > > > > Hi Sohaib, > > > > > > > > > I‘m a student applying for GSOC too. And I've read all of your > discussion > > > in the mail list. > > > > > > I have some questions about your design, and some of the questions may > > > need to be answered by RocketMQ team. So I send them here to be > > discussed. > > > > > > I don't think using key store to persist all the messages is a good > idea. > > > Since MQ is based on O(1) data structure. The key store would harm the > > > performance. > > > > > > I think we can learn from TCP protocol. > > > > > > In Producer-Broker Communication, we can give an incremental id for > every > > > message sent in the same session. And the session id should be > persistent > > > on the disk for producer. So the broker only need to maintain a map > > between > > > session id to expected message id(And this is how Kafka do it). Since > > > messages are much more than producers. However, there's still a K/V > store > > > needed. So we have to ask RocketMQ team about how many producers in the > > > same time while in practical situation. > > > > > > Also, the same idea in Consumer-Broker Communication. > > > > > > > > > About consensus algorithm, I think RocketMQ should already have an > > > implementation there. I don't know what it is, but maybe you can reuse > > > that. Or what if you have to implement one, in my opinion, there's no > > need > > > to implement both Paxos and Raft. Since they solve the same kind of > > > problems. > > > > > > > > > > > > Regards, > > > > > > Dexin > > > > > > > > > > > > 发件人: Sohaib Iftikhar > > > 发送时间: 2018年3月7日 18:15:51 > > > 收件人: dev@rocketmq.apache.org > > > 主题: Re: [GSOC|ROCKETMQ-124] Support non-redundant message delivery > > > mechanism > > > > > > Hi Yukon, > > > > > > Thanks for your reply. Yes, it would be nice to concretely define the > > scope > > > of this project as the doc is a bit ambitious for just a summer. Should > > you > > > (or anyone else) have questions/suggestions/clarifications I'd be glad > > to > > > discuss more details. > > > > > > Thanks, > > > Sohaib > > > > > > On Wed, Mar 7, 2018 at 8:58 AM, yukon wrote: > > > > > > > Hi, > > > > > > > > Google doc is better for discussion, your design is great, now we > could > > > > discuss more details base on it. > > > > > > > > Any advice is welcome from RocketMQ community. > > > > > > > > Appreciate your efforts. > > > > > > > > Regards, > > > > yukon > > > > > > > > On Wed, Mar 7, 2018 at 5:15 AM, Sohaib Iftikhar < > sohaib1...@gmail.com> > > > > wrote: > > > > > > > > > Hi, > > > > > > > > > > @Yukon Thank you for your reply. This clears some doubts. > > > > > > > > > > Sorry for the delay as I was somewhat occupied with another > project. > > I > > > > have > > > > > created an initial design doc. Email is a bit cumbersome for > > feedback I > > > > > wrote this document in two formats: > > > > > > > > > > 1. In the form of a Google document: > > > > > h
