pulsar-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Apache Pulsar Slack" <apache.pulsar.sl...@gmail.com>
Subject Slack digest for #general - 2019-01-29
Date Tue, 29 Jan 2019 09:11:04 GMT
2019-01-28 11:29:39 UTC - Ericko Samudera: Hi @Ezequiel Lovelle, our team is basically trying
to implement chat history pagination. So from the latest message, we take the previous 25
messages, and then the previous 25 messages, and so on. But since this is a chat history,
we're not updating the user's read cursor, just peeking.

We're currently creating a proof-of-concept by modifying pulsar server's `ManagedCursorImpl.asyncGetNthEntry`
to run in the reverse direction, and see where we can go from there.
----
2019-01-28 12:19:10 UTC - Vivek: @Vivek has joined the channel
----
2019-01-28 14:28:43 UTC - jia zhai: @David Tinker If it has no data. use --force is fine.
--force will delete the topic; and producer will reconnect and auto create the topic with
same name.
----
2019-01-28 19:29:28 UTC - Ryan Samo: Hey guys,
Are there any examples of using Spark to Produce data into Pulsar? I see a great example of
consuming messages in the docs. Just wondered if you might have a Spark producer?
----
2019-01-28 19:33:30 UTC - Sijie Guo: Current we only have spark source connector. It should
be easy to add a spark sink connector. If you can create a github issue for it, we will follow
up on implementing one. contributions are welcome as well.
----
2019-01-28 19:34:46 UTC - Ryan Samo: Ok thank you!
ok_hand : Sijie Guo
----
2019-01-28 23:41:45 UTC - Guy Feldman: @Guy Feldman has joined the channel
----
2019-01-29 00:45:55 UTC - Roger Yates: Just to close this off, we resolved this connectivity
issue by the infra guys creating an AWS direct link. Proxy not required!
----
2019-01-29 04:40:08 UTC - Nazrvl: @Nazrvl has joined the channel
----
2019-01-29 06:14:56 UTC - Sarat chandra: @Sarat chandra has joined the channel
----
2019-01-29 07:37:59 UTC - bossbaby: Helo all, I have questions:
If my system is running with the number of zookeeper assuming 3, then I want to add the zookeeper
(scale up), so I have to stop all zookeeper to config again, is that right?
----
2019-01-29 08:15:13 UTC - jia zhai: @bossbaby googled it, this may suit your case :
<https://gist.github.com/miketheman/6057930>
----
2019-01-29 08:44:51 UTC - bossbaby: Is it necessary to stop all bookkeepers and broker to
add zookeeper server?
----
2019-01-29 08:48:43 UTC - jia zhai: @bossbaby It is not necessary to stop bk and broker. since
it is only zk related.
----
2019-01-29 08:53:11 UTC - bossbaby: but bookkeeper vĂ  broker have conf zookeeper in bookie.conf
and broker.conf
----
Mime
View raw message