Re: ceph luminous performance - disks at 100% , low network utilization

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Unfortunately, even after removing all my kernel configuration , the performance did not improve

Currently
GRUB_CMDLINE_LINUX="crashkernel=auto rhgb quiet net.ifnames=0 biosdevname=0 ipv6.disable=1 "

Before
GRUB_CMDLINE_LINUX="crashkernel=auto rhgb quiet net.ifnames=0 biosdevname=0 ipv6.disable=1 intel_pstate=disable intel_idle.max_cstate=0 processor.max_cstate=0 idle=poll numa=off"

This is extremely puzzling - any ideas, suggestions for troubleshooting it will be GREATLY appreciated

Steven

On 2 February 2018 at 10:51, Steven Vacaroaia <stef97@xxxxxxxxx> wrote:
Hi Mark,

Thanks
My pools are using replication =2 

I'll re enable numa and report back

Steven 

On 2 February 2018 at 10:48, Marc Roos <M.Roos@xxxxxxxxxxxxxxxxx> wrote:

Not sure if this info is of any help, please beware I am also just in a
testing phase with ceph.

I don’t know how numa=off is interpreted by the os. If it just hides
the numa, you still could run into the 'known issues'. That is why I
have numad running.
Furthermore I have put an osd 'out' that gives also a 0 in the reweight
column. So I guess your osd.1 is also not participating? If so, could
not be nice if your are testing 3x replication with 2 disks?


I have got this on SATA 5400rpm disks, replicated pool size 3.

rados bench -p rbd 30 write --id rbd
  sec Cur ops   started  finished  avg MB/s  cur MB/s last lat(s)  avg
lat(s)
   20      16       832       816   163.178       180    0.157838
0.387074
   21      16       867       851   162.073       140    0.157289
0.38817
   22      16       900       884   160.705       132    0.224024
0.393674
   23      16       953       937   162.934       212    0.530274
0.388189
   24      16       989       973   162.144       144    0.209806
0.389644
   25      16      1028      1012   161.898       156    0.118438
0.391057
   26      16      1067      1051    161.67       156    0.248463
0.38977
   27      16      1112      1096   162.348       180    0.754184
0.392159
   28      16      1143      1127   160.977       124    0.439342
0.393641
   29      16      1185      1169   161.219       168   0.0801006
0.393004
   30      16      1221      1205   160.644       144    0.224278
0.39363
Total time run:         30.339270
Total writes made:      1222
Write size:             4194304
Object size:            4194304
Bandwidth (MB/sec):     161.111
Stddev Bandwidth:       24.6819
Max bandwidth (MB/sec): 212
Min bandwidth (MB/sec): 120
Average IOPS:           40
Stddev IOPS:            6
Max IOPS:               53
Min IOPS:               30
Average Latency(s):     0.396239
Stddev Latency(s):      0.249998
Max latency(s):         1.29482
Min latency(s):         0.06875


-----Original Message-----
From: Steven Vacaroaia [mailto:stef97@xxxxxxxxx]
Sent: vrijdag 2 februari 2018 15:25
To: ceph-users
Subject: ceph luminous performance - disks at 100% , low
network utilization

Hi,

I have been struggling to get my test cluster to behave ( from a
performance perspective)
Dell R620, 64 GB RAM, 1 CPU, numa=off , PERC H710, Raid0, Enterprise 10K
disks

No SSD - just plain HDD

Local tests ( dd, hdparm ) confirm my disks are capable of delivering
200 MBs
Fio  with 15 jobs indicate 100 MBs
Ceph tell shows  400MBs

rados bench with 1 thread provide  3 MB
rados bench with 32 threads, 2 OSDs ( one per server) , barely touch 10
MB
Adding a third server / OSD improve performance slightly ( 11 MB)

atop shows disk usage at 100% for extended period of time
Network usage is very low
Nothing else is "red"

I have removed all TCP setting  and left ceph.conf mostly with defaults

What am I missing ?

Many thanks

Steven


ceph osd tree
ID  CLASS WEIGHT  TYPE NAME          STATUS REWEIGHT PRI-AFF

  0   hdd 0.54529         osd.0          up  1.00000 1.00000
 -5       0.54529     host osd02
  1   hdd 0.54529         osd.1          up        0 1.00000
 -7             0     host osd04
-17       0.54529     host osd05
  2   hdd 0.54529         osd.2          up  1.00000 1.00000

[root@osd01 ~]# ceph tell osd.0 bench
{
    "bytes_written": 1073741824,
    "blocksize": 4194304,
    "bytes_per_sec": 452125657
}

[root@osd01 ~]# ceph tell osd.2 bench
{
    "bytes_written": 1073741824,
    "blocksize": 4194304,
    "bytes_per_sec": 340553488
}


hdparm -tT /dev/sdc

/dev/sdc:
 Timing cached reads:   5874 MB in  1.99 seconds = 2948.51 MB/sec
 Timing buffered disk reads: 596 MB in  3.01 seconds = 198.17 MB/sec

 fio --filename=/dev/sdc --direct=1 --sync=1 --rw=write --bs=4k
--numjobs=15 --iodepth=1 --runtime=60 --time_based --group_reporting
--name=journal-test
journal-test: (g=0): rw=write, bs=4K-4K/4K-4K/4K-4K, ioengine=sync,
iodepth=1
...
fio-2.2.8
Starting 15 processes
Jobs: 15 (f=15): [W(15)] [100.0% done] [0KB/104.9MB/0KB /s] [0/26.9K/0
iops] [eta 00m:00s]


fio --filename=/dev/sdc --direct=1 --sync=1 --rw=write --bs=4k
--numjobs=5 --iodepth=1 --runtime=60 --time_based --group_reporting
--name=journal-test
journal-test: (g=0): rw=write, bs=4K-4K/4K-4K/4K-4K, ioengine=sync,
iodepth=1
...
fio-2.2.8
Starting 5 processes
Jobs: 5 (f=5): [W(5)] [100.0% done] [0KB/83004KB/0KB /s] [0/20.8K/0
iops] [eta 00m:00s]




_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com


_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux