2020-07-02 09:19:42 UTC - Sijie Guo: @Sijie Guo set the channel topic: - Pulsar 
2.6 is released
<https://pulsar.apache.org/blog/2020/06/18/Apache-Pulsar-2-6-0/>
- Pulsar Summit Talk recordings: 
<https://www.youtube.com/watch?v=wBpNcESySNc&amp;list=PLqRma1oIkcWjVlPfaWlf3VO9W-XWsF_4->
----
2020-07-02 10:58:59 UTC - Joshua Decosta: It seemed like it wasn’t connecting 
to each other. In general, the docs don’t mention how to configure the local 
standalone conf for any of the solutions provided as examples. It would be 
great to know which conf files ply a role when in configuring the standalone 
options as well. Is the standalone conf the only file that needs to be 
configure for tls when using standalone mode? 
----
2020-07-02 11:02:21 UTC - Joshua Decosta: I would also say that the std out 
seems to say that I’m not using tls but when i use the client or admin cli on 
the https routes it works fine. I’m just trying to make sure I’ve configured 
this correctly 
----
2020-07-02 12:11:26 UTC - Durgarao: @Durgarao has joined the channel
----
2020-07-02 12:32:40 UTC - CodeglZhang: @CodeglZhang has joined the channel
----
2020-07-02 17:27:14 UTC - Aaron: I am running into a problem where the pulsar 
admin close function in Java does not clean up the async http threads and the 
delayer thread. Is this a known issue or is there a way around it so that all 
of those resources are actually cleaned up?
----
2020-07-02 17:37:29 UTC - Sijie Guo: Are you using asyncClose or close? It 
should clean up the resources. Otherwise it sounds like a bug.
----
2020-07-02 17:39:10 UTC - Aaron: Close
----
2020-07-02 17:39:56 UTC - Aaron: The delayer thread remains in a TIMED_WAITING 
state
----
2020-07-02 18:01:36 UTC - Sijie Guo: I see. this sounds like a bug. Do you mind 
creating an issue for us?
----
2020-07-02 19:01:26 UTC - Nicolas Ha: Adding 
<https://mvnrepository.com/artifact/org.apache.pulsar/pulsar-client-admin> as a 
dependency to a project makes it fail to load protobuf v3 generated Java 
classes. I see there is a 
<https://mvnrepository.com/artifact/org.apache.pulsar/protobuf-shaded|protobuf-shaded>
 depdendency, which may be related. How can I work around that?
----
2020-07-02 19:04:05 UTC - Aaron: Yes
----
2020-07-02 19:44:50 UTC - Nicolas Ha: If I switch to pulsar-admin-original I 
don’t get the error but instead I get
```         java.lang.NoSuchMethodError: 
org.apache.pulsar.common.util.ObjectMapperFactory.create()Lcom/fasterxml/jackson/databind/ObjectMapper;
org.glassfish.hk2.api.MultiException: A MultiException has 2 exceptions.  They 
are:
                                      1. java.lang.NoSuchMethodError: 
org.apache.pulsar.common.util.ObjectMapperFactory.create()Lcom/fasterxml/jackson/databind/ObjectMapper;
                                      2. java.lang.IllegalStateException: 
Unable to perform operation: create on 
org.apache.pulsar.client.admin.internal.JacksonConfigurator```

----
2020-07-02 19:48:24 UTC - vali: @vali has joined the channel
----
2020-07-02 20:03:15 UTC - vali: @Julius S - would you please share the info you 
have on this with me as well? Thanks!
----
2020-07-02 20:23:49 UTC - Nicolas Ha: tried excluding
```com.fasterxml.jackson.jaxrs jackson-jaxrs-json-provider```
as in this issue without success <https://github.com/apache/pulsar/issues/1409>
----
2020-07-02 20:49:26 UTC - Alexander Ursu: Hi, I was wondering if there is a 
description anywhere of the output of `pulsar-perf monitor-brokers`. I'm not 
sure what is meant by `LONG/S` and `MAX %`.
I am also monitoring the same cluster through it's Prometheus metrics and 
visualising them in Grafana, and I'm noticing that some of the figures don't 
match, specifically for the total `MSG/S` and `KB/S` across the whole cluster.
Does anyone know which is more accurate, or what the differences there are 
between the two?
----
2020-07-02 21:26:49 UTC - Addison Higham: Pulsar brokers write data about their 
load into zookeeper periodically, this data is used for deciding when pulsar 
should load-balance groups of topics across the cluster. `pulsar-perf 
monitor-brokers` ties directly into the data in zookeeper. As far as which 
should be more accurate, that might be a bit hard to say? The load balance data 
is only written periodically and the monitor only grabs the data periodically, 
so it might be "stale" but it should be quite accurate. The prometheus/grafana 
data may not be as accurate (not completely sure of reasons but have heard that 
) but it is emitted more often.

Doing `pulsar-admin topics stats` should be accurate and is also a snapshot 
taken when you ask for it
----
2020-07-02 21:47:17 UTC - Nicolas Ha: I might have solved it with more 
exclusions - need to test the rest of the app to make sure it doesn’t break 
anything
----
2020-07-02 21:51:49 UTC - Oleg Kozlov: Hello, i have a question about metric 
"pulsar_storage_backlog_size" from 
<https://pulsar.apache.org/docs/en/reference-metrics/> page - it says it's a 
number of messages on the doc page, but it seems like it might actually be size 
in bytes. We have a topic with  ~1.6mil delayed messages, and stat 
"storageSize" is ~212mil, and the value of pulsar_storage_backlog_size is also 
close to 212mil... so wanted to check if that's a potentially a typo on the 
reference-metrics page
----
2020-07-02 22:09:32 UTC - Dale Jung: @Dale Jung has joined the channel
----
2020-07-02 22:17:08 UTC - Julius S: Sent you @vali 
----
2020-07-02 23:05:02 UTC - Sijie Guo: good catch. it is a typo
----
2020-07-02 23:05:17 UTC - Sijie Guo: Are you interested in a PR to fix it?
----
2020-07-02 23:06:34 UTC - Oleg Kozlov: yes, thanks for confirming, i can take a 
look at some point :)
----
2020-07-03 02:05:19 UTC - Joshua Eric: @Matteo Merli Can you point me to where 
I might be able to add this functionality? Would be happy to work on it and 
commit back to the repo with a few pointers. Thanks!
----
2020-07-03 04:23:21 UTC - csthomas1311: @csthomas1311 has joined the channel
----
2020-07-03 05:01:43 UTC - Sijie Guo: :+1:
----
2020-07-03 06:21:05 UTC - Matteo Merli: We need to have a way in the Python 
function to allow users to declare the schema of the input event. 

One possible way could be to use an annotation. 

Something like:

```
@JSONSchema(MyRecord)
def myFunc(input):
   ...

```
----
2020-07-03 08:31:42 UTC - Nhat Ha Trinh: @Sijie Guo, Do you have any idea about 
this issue?
----
2020-07-03 08:40:26 UTC - Nhat Ha Trinh: There is no issue with Pulsar 
Encryption when i revert the pulsar version to 2.5.0
Here are my dependencies for 2.5.0
```&lt;dependency&gt;
    &lt;groupId&gt;org.apache.pulsar&lt;/groupId&gt;
    &lt;artifactId&gt;pulsar-client&lt;/artifactId&gt;
    &lt;version&gt;2.5.0&lt;/version&gt;
&lt;/dependency&gt;
&lt;dependency&gt;      &lt;groupId&gt;org.bouncycastle&lt;/groupId&gt;
            &lt;artifactId&gt;bcprov-jdk15on&lt;/artifactId&gt;             
&lt;version&gt;1.60&lt;/version&gt;
&lt;/dependency&gt;```
----

Reply via email to