Hello,
* 2 x Xeon Silver 4212 (12C/24T)
>
I would choose single cpu AMD EPYC systems for lower price with better
performance. Supermicro does have some good systems for AMD as well.
* 16 x 10 TB nearline SAS HDD (8 bays for future needs)
>
Don't waste money here as well. No real gain.
Also, I do not think we saw if the ceph-iscsi versions are the same. We
saw the ceph versions are different which is ok. If you were using a old
ceph-iscsi version which has different settings/values with a newer
version then you could see this error.
On 12/03/2019 06:47 PM, Jason Dillaman wrote:
FYI for ZFS on RBD.
https://github.com/zfsonlinux/zfs/issues/3324
We go for a more modest setting with async to 64, not 2.
--
Alex Gorbachev
Intelligent Systems Services Inc.
On Tue, Dec 3, 2019 at 3:07 PM Fabien Sirjean wrote:
> Hi Ceph users !
>
> After years of using Ceph, we plan to
> Cost of SSD vs. HDD is still in the 6:1 favor of HHD's.
It is not, you can buy fewer thing for less money, with HDDs -that is true
$/TB is better from spinning than from flash, but this is not the most
important indicator, and by far: $/IOPS is another story indeed
On 12/3/19 9:46 PM,
Thanks. If I am reading this correctly the ability to remove an iSCSI
gateway would allow the remaining iSCSI gateways to take over for the
removed gateway's LUN's as of > 3.0. Thats good, we run 3.2. However,
because the actual update of the central config object happens from the
to-be-deleted
It's pretty pointless to discuss erasure coding vs replicated without
knowing how it'll be used.
There are setups where erasure coding is faster than replicated. You
do need to write less data overall, so if that's your bottleneck then
erasure coding will be faster.
Paul
--
Paul Emmerich
> If k=8,m=3 is too slow on HDDs, so you need replica 3 and SSD DB/WAL,
> vs EC 8,3 on SSD, then that's (1/3) / (8/11) = 0.45 multiplier on the
> SSD space required vs HDDs.
> That brings it from 6x to 2.7x. Then you have the benefit of not
> needing separate SSDs for DB/WAL both in hardware cost
If k=8,m=3 is too slow on HDDs, so you need replica 3 and SSD DB/WAL,
vs EC 8,3 on SSD, then that's (1/3) / (8/11) = 0.45 multiplier on the
SSD space required vs HDDs.
That brings it from 6x to 2.7x. Then you have the benefit of not
needing separate SSDs for DB/WAL both in hardware cost and
On Mon, Dec 2, 2019 at 3:42 PM Romit Misra wrote:
>
> Hi Robert,
> I am not quite sure if I get your question correct, but what I understand
> is that you want the inbound writes to land on the cache tier, which
> presumably would be on a faster media, possibily a ssd.
>
> From there you would
>> * Hardware raid with Battery Backed write-cache - will allow OSD to ack
>> writes before hitting spinning rust.
>
> Disagree. See my litany from a few months ago. Use a plain, IT-mode HBA.
> Take the $$ you save and put it toward building your cluster out of SSDs
> instead of HDDs. That way
Hi,
You will get slow performance: EC is slow, HDD are slow too
With 400 iops per device, you get 89600 iops for the whole cluster, raw
With 8+3EC, each logical write is mapped to 11 physical writes
You get only 8145 write IOPS (is my math correct ?), which I find very
low for a PB storage
So,
Rather than a cache tier, I would put an NVMe device in each OSD box
for Bluestore's DB and WAL. This will significantly improve small IOs.
14*16 HDDs / 11 chunks = 20 HDD's worth of write IOPs. If you expect
these files to be written sequentially, this is probably ok.
Mons and mgr on OSD nodes
> After years of using Ceph, we plan to build soon a new cluster bigger than
> what
> we've done in the past. As the project is still in reflection, I'd like to
> have your thoughts on our planned design : any feedback is welcome :)
>
>
> ## Requirements
>
> * ~1 PB usable space for file storage,
Hi Ceph users !
After years of using Ceph, we plan to build soon a new cluster bigger than what
we've done in the past. As the project is still in reflection, I'd like to
have your thoughts on our planned design : any feedback is welcome :)
## Requirements
* ~1 PB usable space for file
I do not think it's going to do what you want when the node you want to
delete is down.
It looks like we only temporarily stop the gw from being exported. It
does not update the gateway.cfg, because we do the config removal call
on the node we want to delete.
So gwcli would report success and
Gateway removal is indeed supported since ceph-iscsi 3.0 (or was it
2.7?) and it works while it is offline :)
Paul
--
Paul Emmerich
Looking for help with your Ceph cluster? Contact us at https://croit.io
croit GmbH
Freseniusstr. 31h
81247 München
www.croit.io
Tel: +49 89 1896585 90
On Tue,
If I recall correctly, the recent ceph-iscsi release supports the
removal of a gateway via the "gwcli". I think the Ceph dashboard can
do that as well.
On Tue, Dec 3, 2019 at 1:59 PM Wesley Dillingham wrote:
>
> We utilize 4 iSCSI gateways in a cluster and have noticed the following
> during
We utilize 4 iSCSI gateways in a cluster and have noticed the following
during patching cycles when we sequentially reboot single iSCSI-gateways:
"gwcli" often hangs on the still-up iSCSI GWs but sometimes still functions
and gives the message:
"1 gateway is inaccessible - updates will be
On Tue, Dec 3, 2019 at 12:58 PM Gesiel Galvão Bernardes
wrote:
>>
>> Jason,
>
>
> Returned a json with "data" and a different strnig. Were they supposed to be
> the same?
Yes, they should be the same since those are the hashes of the config.
> Now I checked that ceph version is different in
> Jason,
Returned a json with "data" and a different strnig. Were they supposed to
be the same?
Now I checked that ceph version is different in the gateways: ceph-iscsi1
running 13.2.6 and ceph-iscsi1 running 13.2.7. Could this be the problem?
Gesiel
Em ter., 3 de dez. de 2019 às 14:43, Jason
You might need to reset the target API service. You can compare what
the two REST APIs think the config is by running:
$ curl --insecure --user : -X GET
http://ceph-iscsi1:5000/api/sysinfo/checkconf
$ curl --insecure --user : -X GET
http://ceph-iscsi3:5000/api/sysinfo/checkconf
(tweak username,
Are exactly the same:
[root@ceph-iscsi1 ~]# sha256sum /etc/ceph/iscsi-gateway.cfg
33620867c9c3c5e6a666df2bb461150d4a885db2989cd3aea812a29555fdc45a
/etc/ceph/iscsi-gateway.cfg
[root@ceph-iscsi3 ~]# sha256sum /etc/ceph/iscsi-gateway.cfg
The sha256sum between the "/etc/ceph/iscsi-gateway.cfg" where you are
running the 'gwcli' tool and `ceph-iscsi3` most likely mismatches.
On Tue, Dec 3, 2019 at 12:19 PM Gesiel Galvão Bernardes
wrote:
>
> Hi everyone,
>
> I have a problem trying to add an ISCSI gateway. The following error is
>
On Tue, Nov 26, 2019 at 7:45 PM majia xiao wrote:
> Hi all,
>
> We have a Ceph(version 12.2.4)cluster that adopts EC pools, and it
> consists of 10 hosts for OSDs.
>
> The corresponding commands to create the EC pool are listed as follows:
>
>
>
> ceph osd erasure-code-profile set
According to my understanding, osd's heartbeat partners only come from
those osds who assume the same pg
Hello,
That was my initial assumption too.
But according to my experience set of heartbeat peers include pg peers
and some other osds.
Actually it contains:
- pg peers
- next and
I created one ceph cluster.
node-1: mon, mgr, osd.0, mds
node-2: mon, mgr, osd.1, mds
node-3: mon, mgr, osd.2, mds
When the cluster is working normally, using command "mount -t ceph
:/ /mnt -o name=admin,secret=" to mount is
ok.
But when a node unusual down(like poweroff), and using same
Hi,
I have set
upmap_max_iterations 2
w/o any impact.
In my opinion the issue is that the evaluation of OSDs data load is not
working.
Or can you explain why osdmaptool does not report anything to do?
Regards
Thomas
Am 03.12.2019 um 08:26 schrieb Harald Staub:
> Hi all
>
> Something to try:
>
27 matches
Mail list logo