Could you attach something simple as VisualVM, https://visualvm.java.net/
to the process when it runs CC and sample on memory to get an idea of what
it is that takes up so much memory?
On Mon, Mar 6, 2017 at 1:35 PM, unrealadmin23 via Neo4j <
neo4j@googlegroups.com> wrote:
> As an update - after
As an update - after running for 48 hours the GC again runs out of heap. I
can increase the heap size again (above 124GB) but getting a bit tedious
now.
There are no indicators that the DB has a problem - I was just interested
to ensure that its integrity was good.
Let me know if you wish
Looks like this fixed the problem - for the Heap size.
Setting Env variables doesn't appear to work - will you be coding this in
to the script ?
In my case the 90% phase comprises a busy machine, with little IO
activity. The data appears to have been loaded into the memory and the
So I have added:
$ cat neo4j-admin
#!/usr/bin/env bash
# Copyright (c) 2016 "Neo Technology,"
# Network Engine for Objects in Lund AB [http://neotechnology.com]
#
# This file is part of Neo4j.
#
# Neo4j is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero
Can you try to edit the script directly and add the memory parameters there?
On Fri, Mar 3, 2017 at 8:49 PM, unrealadmin23 via Neo4j <
neo4j@googlegroups.com> wrote:
> Yes
>
> Also, in the 90% scan, what every Java memory parameter I use, htop
> shows the same memory foot print. Its as if
Yes
Also, in the 90% scan, what every Java memory parameter I use, htop shows
the same memory foot print. Its as if the heap isn't being set as per the
env parameters that you area asking me to set.
Wayne
On Friday, 3 March 2017 07:51:48 UTC, Mattias Persson wrote:
>
> Querying Lucene,
Sorry I just learned that neo4j-admin uses a different variable
"You can pass memory options to the JVM via the `JAVA_MEMORY_OPTS` variable as
a workaround though."
Von meinem iPhone gesendet
> Am 28.02.2017 um 18:50 schrieb unrealadmin23 via Neo4j
> :
>
> Michael,
Michael,
After running the check_consistency command for 1 day with the above
parameters, it failed in exactly the same manner.
$env | grep -i java
JAVA_OPTS=-Xmx32G -Xms32G
Any other ideas ?
Wayne
On Monday, 27 February 2017 16:57:49 UTC, Michael Hunger wrote:
>
> Do you have really that
Also if you use Neo4j Enterprise with a contract you can also raise support
issues with Zendesk :)
On Mon, Feb 27, 2017 at 8:47 PM, unrealadmin23 via Neo4j <
neo4j@googlegroups.com> wrote:
> I have 1TB so usual consider 120GB to be small; but then I'm not a Java
> person.
> I will set the
I have 1TB so usual consider 120GB to be small; but then I'm not a Java
person.
I will set the java_opts as per below and see what happens
Wayne.
On Monday, 27 February 2017 16:57:49 UTC, Michael Hunger wrote:
>
> Do you have really that much RAM in your machine ? 120G usually doesn't
>
Do you have really that much RAM in your machine ? 120G usually doesn't
make sense. Most people run with 32G as large heap.
That said. I asked and currently the numbers from the config are not used,
you have to do:
export JAVA_OPTS=-Xmx24G -Xms24G
neo4j-admin ...
On Mon, Feb 27, 2017 at 8:32
I should have said, that the head sizes are the ones that I have set in
neo4j.conf.
Will these be used by check-consistency or do I need to supply them
elsewhere ?
Wayne.
On Monday, 27 February 2017 07:27:33 UTC, unreal...@googlemail.com wrote:
>
> Michael,
>
> neo4j-admin check-consistency
Michael,
neo4j-admin check-consistency --database=test.db --verbose
dbms.memory.heap.initial_size=12m
dbms.memory.heap.max_size=12m
Wayne.
On Monday, 27 February 2017 02:47:26 UTC, Michael Hunger wrote:
>
> How did you call the consistency checker?
>
> How much heap did you provide
How did you call the consistency checker?
How much heap did you provide for it?
Cheers, Michael
On Sun, Feb 26, 2017 at 8:28 PM, unrealadmin23 via Neo4j <
neo4j@googlegroups.com> wrote:
> The following o/p was obtained:
>
> .
> .
> .
>
> 90%
> 2017-02-26
The following o/p was obtained:
.
.
.
90%
2017-02-26 00:03:16.883+ INFO [o.n.c.ConsistencyCheckService] ===
Stage7_RS_Backward ===
2017-02-26 00:03:16.885+ INFO [o.n.c.ConsistencyCheckService] I/Os
RelationshipStore
Reads: 3374851294
Random Reads: 2743390177
15 matches
Mail list logo