Bug 996746 - Reading from glusterfs drive takes twice as much data transfer as what you read
Summary: Reading from glusterfs drive takes twice as much data transfer as what you read
Keywords:
Status: CLOSED EOL
Alias: None
Product: GlusterFS
Classification: Community
Component: unclassified
Version: pre-release
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: bugs@gluster.org
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2013-08-13 21:05 UTC by Federico Leva
Modified: 2015-10-22 15:40 UTC (History)
2 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2015-10-22 15:40:20 UTC
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Embargoed:


Attachments (Terms of Use)
Network graph of an instance after starting upload of some big files (15.47 KB, image/png)
2013-09-09 06:26 UTC, Federico Leva
no flags Details

Description Federico Leva 2013-08-13 21:05:59 UTC
Description of problem:
Reading from glusterfs is too slow and is easily capped; network consumption is twice as needed.

Version-Release number of selected component (if applicable):
$ glusterfs -V
glusterfs 3.3.1 built on Feb 14 2013 17:36:30
$ uname -a
Linux i-00000355 3.2.0-48-virtual #74-Ubuntu SMP Thu Jun 6 20:02:55 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

How reproducible:
Always

Steps to Reproduce:
1. Put some file on a glusterfs drive
2. Read the file and upload it somewhere else
3. Monitor the network stats

Actual results:
The download speed from glusterfs is exactly twice as the upload speed i.e. read speed, always and constantly. Typical situation:

NetHogs version 0.8.0

  PID USER     PROGRAM                                                                                                                       DEV        SENT      RECEIVED       
8151  root     /usr/sbin/glusterfs                                                                                                           eth0      36.996    7022.670 KB/sec
7417  nemobis  perl                                                                                                                          eth0    3364.814      21.309 KB/sec
?     root     10.4.0.218:47259-208.80.152.32:389                                                                                                       0.000       0.000 KB/sec
19530 root     /usr/sbin/glusterfs                                                                                                           eth0       0.000       0.000 KB/sec
11766 nemobis  sshd: nemobis@pts/0                                                                                                           eth0       0.000       0.000 KB/sec
?     root     unknown TCP                                                                                                                              0.000       0.000 KB/sec

  TOTAL                                                                                                                                              3401.810    7043.980 KB/sec 






Expected results:
Glusterfs sends me only the data I need, not wasting capacity over mysterious duplication of data transfer.

Additional info:
Tested from an instanceof wmflabs.org / Wikimedia Labs; I've noticed this for a long time now, also very easily seen on ganglia graphs.

Comment 1 Federico Leva 2013-08-13 21:07:05 UTC
Too much whitespace...

NetHogs version 0.8.0

  PID USER     PROGRAM               DEV        SENT      RECEIVED       
8151  root     /usr/sbin/glusterfs   eth0      77.421   15704.657 KB/sec
7417  nemobis  perl                  eth0    7800.735      60.180 KB/sec
1097  nslcd    /usr/sbin/nslcd       eth0       0.000       0.000 KB/sec
19530 root     /usr/sbin/glusterfs   eth0       0.000       0.000 KB/sec
11766 nemobis  sshd: nemobis@pts/0   eth0       0.000       0.000 KB/sec
?     root     unknown TCP                      0.000       0.000 KB/sec

  TOTAL                                      7878.156   15764.838 KB/sec

Comment 2 Federico Leva 2013-09-09 06:26:39 UTC
Created attachment 795504 [details]
Network graph of an instance after starting upload of some big files

Comment 3 Kaleb KEITHLEY 2015-10-22 15:40:20 UTC
pre-release version is ambiguous and about to be removed as a choice.

If you believe this is still a bug, please change the status back to NEW and choose the appropriate, applicable version for it.


Note You need to log in before you can comment on or make changes to this bug.