Hello,
Ignite compresses each page individually. The result of whole file
compression will always be better than the result of each individual page
compression. Moreover, Ignite stores compressed pages only if the page size
shrunk by one or more filesystem blocks. So, for example, if you have fs
b
Hello,
This problem was only in Ignite 2.8 with enabled partition awareness. Fixed
in Ignite 2.8.1, see [1].
[1]: https://issues.apache.org/jira/browse/IGNITE-12743
вт, 17 нояб. 2020 г. в 01:53, Hemambara :
> Ignite thin client create and use ReliableChannel which starts below async
> thread, b
Which version are you using?
A similar issue was fixed for 2.8.1 and 2.9:
https://issues.apache.org/jira/browse/IGNITE-12743
On Tue, Nov 17, 2020 at 1:53 AM Hemambara wrote:
> Ignite thin client create and use ReliableChannel which starts below async
> thread, but while closing the channel it is
Hi,
Please find the attached log for a complete but failed reboot. You can see
the exceptions.
On Mon, Nov 16, 2020 at 4:00 AM Ivan Bessonov wrote:
> Hello,
>
> there must be a bug somewhere during node start, it updates its
> distributed metastorage content and tries to join an already activat
I have enabled compression
(pageSize=16384, diskPageCompression=ZSTD, diskPageCompressionLevel=18) but
the partition files don't appear to be very compressed. I tested by adding
approx 16000 data items to my cache and looking at the partition files on
disk.
Example: part-96.bin is 339M in size. If
Ignite thin client create and use ReliableChannel which starts below async
thread, but while closing the channel it is not shutting down this executor
service thread. This should be fine if jvm shutsdown, but it will be a
problem for applications sharing jvm node. Ex: 1 JVM has multiple
application
Can you also provide the logs for the few minutes before the "Node is out of
topology" message?
Igor
--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/
Hi we can't use the thin client unfortunately because we rely heavily on
continuous queries to notify clients when an update has occurred without
delay.
based on the heap dump above, is there anything we can do to get around this
large mem footpring on the client? It seems odd that the client need
Hi Ilya,
Regarding exchange worker, the jira saying that 2.8.1 and 2.9 but we are
using 2.7.6 version. So it could be in 2.7.6 as well. But still I am not
understanding here is, there is no node join/left or cluster activation or
cache creation near to the error logging time. Don't know why PME tr
Hello!
I recommend taking a heap dump from this node when it is nearing OOM,
checking it with e.g. Eclipse MAT to see what is the leak suspect/class
histogram.
Then share the findings with us.
--
Ilya Kasnacheev
ср, 4 нояб. 2020 г. в 19:45, Ravi Makwana :
> HI,
>
> We are using Apache Ignite
Hello!
1. There's no such thing as pure on-heap cache. On-heap is an extra option
for a cache, which is always off-heap. So you need to set both.
2. Eviction policy is applicable to on-heap part of a cache only. Expiry
policy will remove data from both off-heap and PDS. There's also page
eviction
Hello!
For the first 4 thread dump, the problem was in establishing communication
connection to one of nodes:
[2020-11-09T02:31:41,105][WARN
][tcp-comm-worker-#1%EDIFCustomerCC%][TcpCommunicationSpi] Connect timed
out (consider increasing 'failureDetectionTimeout' configuration property)
[addr=/k
Hi All,
I need confirmation for my understanding of some basic concepts.
These are my understanding. Please confirm.
1) Regions are not applicable to on heap caches. I'd use JVM -Xms and -Xmx
to set the limits, while with off heap caches, I'd use regions with
initial/max size set.
2) When I defin
I can Post an detailed Error mag tonight
Sent from Nine
Von: Stephen Darlington
Gesendet: Montag, 16. November 2020 12:10
An: user
Betreff: Re: Thin Client connection not working...
Doesn’t work how? Doesn’t compile? Doesn’t connect? Doesn’t create the cache?
I
Unable to connect
Sent from Nine
Von: Stephen Darlington
Gesendet: Montag, 16. November 2020 12:10
An: user
Betreff: Re: Thin Client connection not working...
Doesn’t work how? Doesn’t compile? Doesn’t connect? Doesn’t create the cache?
Is there an error?
> O
Hello!
I have no idea and I think it depends on the StreamReceived/allowOverwrite.
Please try it and see.
Regards,
--
Ilya Kasnacheev
пт, 13 нояб. 2020 г. в 18:06, ssansoy :
> Last question would this work with datastreamer? e.g. adding a field
> inside the transformation?
> Thanks!
>
>
>
Doesn’t work how? Doesn’t compile? Doesn’t connect? Doesn’t create the cache?
Is there an error?
> On 16 Nov 2020, at 09:57, Wolfgang Meyerle
> wrote:
>
> Hi,
>
> I tried using the cpp thin client example from the Apache ignite site to
> create a small thin client connection example to one
Hi,
I tried using the cpp thin client example from the Apache ignite site to
create a small thin client connection example to one of the running
Apache Ignite cluster nodes.
However it doesn't work and I'm out of a clue.
I added the following bean to my persistence configuration file:
id="i
Hello,
there must be a bug somewhere during node start, it updates its
distributed metastorage content and tries to join an already activated
cluster, thus creating a conflict. It's hard to tell the exact data that
caused conflict, especially without any logs.
Topic that you mentioned (
http://ap
19 matches
Mail list logo
