Can you help with more info here?
The RocksDB backend itself is in ABFS instead of local? Or you mean the
checkpoint is in ABFS but local dir for RocksDB is in local storage?
GetPathSTatus is done by your monitoring pages? We run Flink on ABFS so we
would like to see if we can help you out.
Hi Jose
This component is compatible with the latest version of Flink (1.17.1). The
package in question is merely a shaded dependency for Hadoop and is not
dependent on the version of Flink.
Best,
Feng
On Thu, Sep 28, 2023 at 4:03 AM Jose Henrry Matituy Manchabajoy <
henrrymati...@gmail.com>
Cordial greetings
I am working on an application where I need to use Flik with HDFS. I
understand that to make the connection I need to use the
flink-shaded-hadoop-2-uber-2.8.3-10.0.jar extension. My question is. What
is the most up to date version of Flink that works with this extension or
is
Hello,
We are using ABFSS for RocksDB's backend as well as the storage dir
required for Kubernetes HA. In the Azure Portal's monitoring insights I see
that every single operation contains failing transactions for the
GetPathStatus API. Unfortunately I don't see any additional details, but I
know
hi Feng，
Thanks for your reply, we are 8 u192 JDK, may be is the question, I found a
JDK issue:https://bugs.openjdk.org/browse/JDK-8215355.
Best,
rui
Feng Jin 于2023年9月27日周三 20:09写道：
> hi Rui,
>
> Which version of JDK are you using?
>
> This issue could potentially be a bug in the JDK version.
hi Rui,
Which version of JDK are you using?
This issue could potentially be a bug in the JDK version.
If you are using JDK 8, you can try using OpenJDK 8u265 as a possible
solution.
Best,
Feng
On Wed, Sep 27, 2023 at 8:08 PM rui chen wrote:
>
>
> rui chen 于2023年9月27日周三 19:32写道：
>
>> hi
rui chen 于2023年9月27日周三 19:32写道：
> hi Feng,
>
> Thank you for your reply，We observed the GC situation, there is no change
> before and after replacement, several tasks on our line using jemalloc have
> appeared stuck, after removing jemalloc, no stuck situation has been found.
>
> Best,
> rui
>
>
Have you checked the metric changes in 1.17.
>From release notes 1.17:
https://nightlies.apache.org/flink/flink-docs-master/release-notes/flink-1.17/
Metric Reporters #
Only support reporter factories for instantiation #
FLINK-24235 #
Configuring reporters by their class is no longer supported.
hi rui,
In general, checkpoint timeouts are typically associated with the job's
processing performance. When using jemalloc, performance degradation is
generally not observed.
It is advisable to analyze whether the job's garbage collection (GC) has
become more frequent.
Best,
Feng
On Mon,
Hi Rania
Currently, there are no similar metrics available. Only the number of
memory segments in the queue between tasks
(inputQueueLength/outputQueueLength) and the buffer occupancy ratio
(inPoolUsage/outPoolUsage) are available.
For more detailed information, please refer to:
Some more details on my problem:
1. The "Multiple implementations" problem was because I had the
metrics-prometheus jar both in the plugins and lib directories. I
tried putting it in only one,
and in both cases (plugins or lib), the result was the same, I got
only Flink metrics on my prom port.
11 matches
Mail list logo
