[ceph-users] RadosGW performance s3 many objects

Hi,we are using the Ceph with RadosGW and S3 setting.With more and more objects in the storage the writing speed slows down significantly. With 5 million object in the storage we had a writing speed of 10MS/s. With 10 million objects in the storage its only 5MB/s. Is this a common issue?Is the

Re: [ceph-users] Ceph OSD network configuration

Hi, you can/need to configure the clusternetwork and the publicnetwork. In the ceph.conf that means: public_network = 1.2.3.0/24 cluster_network = 2.3.4.0/24 as an example. So you will, for every node in your ceph cluster, use 1x 10G Port for public network ip 1x 10G Port for cluster network

[ceph-users] Performance - pool with erasure/replicated type pool

Hello, I have 2 pools as below: 1. pool1 with erasure type 2. pool2 with replicated type I ran the "rados bench" with above 2 pool and the results came as below: - Read performance - around 60% better for replicated type pool ie pool2 - Write performance - around 50 % better for erasure type poo

Re: [ceph-users] Performance - pool with erasure/replicated type pool

Yes that seems about right. Erasure coding works similar to Raid5/6 so data will be striped, whereas replicated pools are simple copies of data. When you write to a 3x replicated pool you are having to write 3 times as much and so performance is lower. When writing to an erasure coded pool (k=8 m=

Re: [ceph-users] Ceph OSD network configuration

If you want redundancy and link aggregation then you will use LACP (or one of the Linux pseudo modes) and bonding in linux. This is common and not Ceph specific. Robert LeBlanc Sent from a mobile device please excuse any typos. On Jan 23, 2016 6:39 PM, "名花" wrote: > > Hi, I have a 4 ports 10gb

Re: [ceph-users] Performance - pool with erasure/replicated type pool

Thanks for details. What is the best suitable pool type for object storage? Thanks Swami On Sun, Jan 24, 2016 at 8:58 PM, Nick Fisk wrote: > Yes that seems about right. > > Erasure coding works similar to Raid5/6 so data will be striped, whereas > replicated pools are simple copies of data. When

Re: [ceph-users] move/upgrade from straw to straw2

Did it work for you to just change 'straw' to 'straw2' in your crushmap? Thanks, Shinobu On Mon, Sep 21, 2015 at 9:39 PM, Stefan Priebe - Profihost AG < s.pri...@profihost.ag> wrote: > > Am 21.09.2015 um 13:47 schrieb Wido den Hollander: > > > > > > On 21-09-15 13:18, Dan van der Ster wrote: > >

[ceph-users] About mon_osd_full_ratio

Hi All, I use 'ceph tell' command to inject an argument mon_osd_full_ratio by a certain value like 10%, finding 'Rados bench' can write more than 10%(it may slow down when approaching 10%, you can interrupt the process and restart executing with 'rados bench'). Only 'ceph pg set_full_ratio

[ceph-users] Unable to delete file in CephFS

Hi, there's a rogue file in our CephFS that we are unable to remove. Access to the file (removal, move, copy, open etc.) results in the MDS starting to spill out the following message to its log file: 2016-01-25 08:39:09.623398 7f472a0ee700 0 mds.0.cache open_remote_dentry_finish bad remote

[ceph-users] optimized SSD settings for hammer

Hi, is there a guide or recommendation to optimized SSD settings for hammer? We have: CPU E5-1650 v3 @ 3.50GHz (12 core incl. HT) 10x SSD / Node journal and fs on the same ssd currently we're runnig: - with auth disabled - all debug settings to 0 and ms_nocrc = true osd_op_num_threads_per_shar

Re: [ceph-users] optimized SSD settings for hammer

ms_nocrc options is changed to the following in Hammer.. ms_crc_data = false ms_crc_header = false Rest looks good , you need to tweak the shard/thread based on your cpu complex and total number of OSDs running on a box.. BTW, with latest Intel instruction sets crc overhead is re