Hi all,
I have a cluster in jewel which is upgraded from hammer , I want to migrate
ms_type to async, but I can't find documents about it. Does somebody know how
to do that?
Thanks
Choury
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://li
You just have the set the config option ("ms_type = async", I think?) and
restart daemons. Both messengers use the same protocol and there's no
migration in terms of data or setting up compatibility.
-Greg
On Thu, Jan 25, 2018 at 9:43 AM 周 威 wrote:
> Hi all,
>
> I have a cluster in jewel which i
Hi Greg,
Can I restart daemons one by one online? Or there will be a down time before I
restart all daemons?
Thanks
Choury
发件人: Gregory Farnum
发送时间: 2018年1月25日 16:52:46
收件人: 周 威
抄送: ceph-users@lists.ceph.com
主题: Re: [ceph-users] How to migrate ms_type
Yes, you can (and should!) restart them one at a time to avoid downtime.
On Thu, Jan 25, 2018 at 9:57 AM 周 威 wrote:
> Hi Greg,
>
> Can I restart daemons one by one online? Or there will be a down time
> before I restart all daemons?
>
> Thanks
> Choury
>
> __
I’ve test that, and yes. The cluster works fine with some node using
async-message while others using simple-message.
Thank you, Greg.
Yes, you can (and should!) restart them one at a time to avoid downtime.
On Thu, Jan 25, 2018 at 9:57 AM 周 威 mailto:cho...@msn.cn>> wrote:
Hi Greg,
Can I rest
Well I can absolsulty try that or I can even make a mix of SPDK and non SPDK
OSDs and test them individually
I just wanted to know how stable is it and if I could use it as RocksDB
Thanks for the help
Mensaje original De: Igor Fedotov Fecha:
24/1/18 12:26 p. m. (GMT+01:00) Pa
Hey Cephers,
Sorry for the short notice, but the Ceph Tech Talk schedule for today
(January 25th) has been canceled because a calendar conflict.
We will be posting the schedule for the upcoming Ceph Tech Talks here
soon. If you have any interesting topic to talk about, feel free to
contact me.
K
Hey Cephers,
This is a friendly reminder that the Call for Proposals for the
Cephalocon APAC 2018[1] ends next Wednesday, January 31st.
[1] http://cephalocon.doit.com.cn/guestreg_en.html
If you haven't submitted your proposal so far, you still have time!
Kindest regards,
Leo
--
Leonardo Vaz
Hi,
setting the pplication pool helped - the performance is not skewed anymore
( i.e. SSDpool is better than HDD)
However latency when using more threads is still very high
I am getting 9.91 Gbits/sec when testing with iperf
Not sure what else should I check
As always, your help will be greatly a
Hey Eugen,
Pleasure. Glad you problem resolved itself.
Regards,
Tom
On Wed, Jan 24, 2018 at 5:29 PM, Eugen Block wrote:
> Hi Tom,
>
> thanks for the detailed steps.
>
> I think our problem literally vanished. A couple of days after my email I
> noticed that the web interface suddenly listed o
Hello,
We are running Luminous 12.2.2. 6 OSD hosts with 12 1TB OSDs, and 64GB
RAM. Each host has a SSD for Bluestore's block.wal and block.db.
There are 5 monitor nodes as well with 32GB RAM. All servers have
Gentoo with kernel, 4.12.12-gentoo.
When I export an image using:
rbd export pool-name/
Yesterday I noticed some OSDs were missing from our cluster (96 OSDs total,
84up/84in is what showed).
After drilling down to determine which node and the cause, I found that all the
OSDs on that node (12 total) were in fact down.
I entered 'systemctl status ceph-osd@$osd_number' to determine e
On 2018/01/25 2:03 pm, Andre Goree wrote:
Yesterday I noticed some OSDs were missing from our cluster (96 OSDs
total, 84up/84in is what showed).
After drilling down to determine which node and the cause, I found
that all the OSDs on that node (12 total) were in fact down.
I entered 'systemctl s
The first and hopefully easy one:
I have a situation where I have two pools that are rarely used (the third will
be in use after I can get through these issues), but they need to present at
the whims of our cloud team. Is there a way I can turn off ‘2 pools have many
more objects per pg than av
On Wed, Jan 24, 2018 at 7:47 AM, Keane Wolter wrote:
> Hello all,
>
> I was looking at the Client Config Reference page
> (http://docs.ceph.com/docs/master/cephfs/client-config-ref/) and there was
> mention of a flag --client_with_uid. The way I read it is that you can
> specify the UID of a user
On Fri, Jan 26, 2018 at 5:47 AM, Andre Goree wrote:
> On 2018/01/25 2:03 pm, Andre Goree wrote:
>>
>> Yesterday I noticed some OSDs were missing from our cluster (96 OSDs
>> total, 84up/84in is what showed).
>>
>> After drilling down to determine which node and the cause, I found
>> that all the O
The doc at
http://docs.ceph.com/docs/master/ceph-volume/lvm/prepare/#ceph-volume-lvm-prepare
says I can pass a physical device to ceph-volume. But when I try to do that:
gentooserver ~ # ceph-volume lvm create --bluestore --data /dev/sdb
usage: ceph-volume lvm create [-h] [--journal JOURNAL] --dat
Did you wipe all of the existing partitions and such first? Which version
of ceph? The below commands are what I ran to re-add my osds as bluestore
after moving all data off of them.
ceph-volume lvm zap /dev/sdb
ceph-volume lvm create --bluestore --data /dev/sdb
On Thu, Jan 25, 2018 at 9:41 PM
-- Forwarded message --
From: Nathan Dehnel
Date: Thu, Jan 25, 2018 at 9:49 PM
Subject: Re: [ceph-users] ceph-volume raw disks
To: David Turner
>Did you wipe all of the existing partitions and such first?
I tried it both before and after creating an lvm physical partition.
>Whi
Hi,
We have set no scrub , no deep scrub flag on a ceph cluster.
When we are deleting snapshots we are not seeing any change in usage space.
I understand that Ceph OSDs delete data asynchronously, so deleting a
snapshot doesn’t free up the disk space immediately. But we are not seeing
any change
Hi List,
I read a old article about how ceph client read from ceph cluster.It said the
client only read from the primary osd. Since ceph cluster in replicate mode
have serveral copys of data only read from one copy seems waste the performance
of concurrent read from all the copys.
But that artci
Hi Peter,
Not sure if you have got to the bottom of your problem, but I seem to have
found what might be a similar problem. I recommend reading below, as there
could be a potential hidden problem.
Yesterday our cluster went into *HEALTH_WARN* state and I noticed that one
of my pg's was listed a
22 matches
Mail list logo