Ask Your Question
0

does this look right? [closed]

asked 2014-05-26 06:11:04 -0500

blazesurfer gravatar image

My rings

srv:/etc/swift # swift-ring-builder container.builder
account.builder, build version 20
262144 partitions, 1.000000 replicas, 1 regions, 1 zones, 4 devices, 0.73 balance
The minimum number of hours before a partition can be reassigned is 1
Devices:    id  region  zone      ip address  port      name weight partitions balance meta
             0     1     1       127.0.0.1  6012      sdb1   1.00         66    0.73
             1     1     1       127.0.0.1  6012      sdc1   0.00          0    0.00
             2     1     1       127.0.0.1  6012      sdd1 2000.00     131039   -0.00
             3     1     1       127.0.0.1  6012      sde1 2000.00     131039   -0.00
srv:/etc/swift # swift-ring-builder container.builder
container.builder, build version 20
262144 partitions, 1.000000 replicas, 1 regions, 1 zones, 4 devices, 0.79 balance
The minimum number of hours before a partition can be reassigned is 1
Devices:    id  region  zone      ip address  port      name weight partitions balance meta
             0     1     1       127.0.0.1  6011      sdb1   1.00         65   -0.79
             1     1     1       127.0.0.1  6011      sdc1   0.00          0    0.00
             2     1     1       127.0.0.1  6011      sdd1 2000.00     131040    0.00
             3     1     1       127.0.0.1  6011      sde1 2000.00     131039   -0.00
srv:/etc/swift # swift-ring-builder object.builder
object.builder, build version 17
262144 partitions, 1.000000 replicas, 1 regions, 1 zones, 4 devices, 0.79 balance
The minimum number of hours before a partition can be reassigned is 1
Devices:    id  region  zone      ip address  port      name weight partitions balance meta
             0     1     1       127.0.0.1  6010      sdb1   1.00         65   -0.79
             1     1     1       127.0.0.1  6010      sdc1   0.00          0    0.00
             2     1     1       127.0.0.1  6010      sdd1 2000.00     131039   -0.00
             3     1     1       127.0.0.1  6010      sde1 2000.00     131040    0.00

Disk utilization

    srv:/etc/swift # df -h
Filesystem      Size  Used Avail Use% Mounted on
/dev/sda2        28G  3.1G   24G  12% /
udev            1.9G  140K  1.9G   1% /dev
tmpfs           1.9G     0  1.9G   0% /dev/shm
/dev/sdb1       2.0T  2.0T   22G  99% /media/swift/sdb1
/dev/sdc1       2.0T  2.0T  7.6G 100% /media/swift/sdc1
/dev/sdd1       2.0T  2.6G  2.0T   1% /media/swift/sdd1
/dev/sde1       2.0T  2.5G  2.0T   1% /media/swift/sde1

Recon -d

srv:/etc/swift # swift-recon -d
===============================================================================
--> Starting reconnaissance on 1 hosts
===============================================================================
[2014-05-26 21:07:30] Checking disk usage now
Distribution Graph:
  0%    2 *********************************************************************
 98%    1 **********************************
 99%    1 **********************************
Disk usage: space used: 4370558619648 of 8791793860608
Disk usage: space free: 4421235240960 of 8791793860608
Disk usage: lowest: 0.12%, highest: 99.63%, avg: 49.7117958967%
===============================================================================

Doesn't seem to be balancing.. and cant seem to get this to work as needed to level the disks out, I want to expand to multi node and multi replica though need to get the disks level'd first as it ... (more)

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by SamYaple
close date 2014-06-06 13:00:53.082030

1 answer

Sort by ยป oldest newest most voted
1

answered 2014-05-28 09:06:07 -0500

SamYaple gravatar image

updated 2014-05-28 09:07:55 -0500

Look at your "weight" parameter. You have set it to 2000 for the new drives but for the original members it is set to 1 or 0. This will screwup the balance. You can change the weight for those original ones and run the rebalance command and you should be set.

# swift-ring-builder object.builder set_weight z1-127.0.0.1:6010/sdb1 2000 //Change weight
# swift-ring-builder object.builder                                        //Verify the changes
# swift-ring-builder object.builder rebalance                              //Commit the changes

You may also want change your ip from 127.0.0.1 to your actual ip address as you will have to do that to have multiple nodes.

edit flag offensive delete link more

Comments

Thank you for your response,

can i do the weight and ip change at same time? Also that was to try and evacuate those two drives. as it wasnt balancing properly. so ill change back to 2000 on all drives.

ill need to remove each drive and re add with the correct ip yes? as well at the same time?

blazesurfer gravatar imageblazesurfer ( 2014-06-05 23:29:44 -0500 )edit

In this case, the set_weight function will only change the weight. It uses z1-127.0.0.1:6010/sdb1 as the device to modify. It doesn't allow you to change that device to a different one (change ip address). But there my be a function out there that can do that. I am unsure.

Do some research and try it is my best advice.

SamYaple gravatar imageSamYaple ( 2014-06-06 08:08:16 -0500 )edit

Thanls SamYaple

It all appears well now as per previous post. just watching logs then will add second node.

blazesurfer gravatar imageblazesurfer ( 2014-06-06 08:11:31 -0500 )edit

Great! I just reconized your name from the other post. I didn't realize I was responding to the same person ;) Let me know if you have anymore issue!

SamYaple gravatar imageSamYaple ( 2014-06-06 13:00:46 -0500 )edit

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2014-05-26 06:11:04 -0500

Seen: 144 times

Last updated: May 28 '14