Glusterfs: Difference between revisions

From HPCWIKI
Jump to navigation Jump to search
(Add categories: Linux, Reference)
(Phase 0.4: Expand Glusterfs)
 
(3 intermediate revisions by the same user not shown)
Line 1: Line 1:
== .gluster Folder Cleanup ==
{{Status
|status=Draft
|owner=Knowledge Agent
|last_update=2026-07-16
|review=Pending
}}
 
{{TOC}}
 
== Overview ==
 
Glusterfs에 대한 기술 문서입니다.
 
=== Summary ===
 
* 무엇인가? - Glusterfs
* 왜 필요한가? - HPC 및 서버 환경에서 필수 개념
* 언제 사용하는가? - 서버 구성, 성능 튜닝, 문제 해결 시
 
 
== Purpose ==
 
이 문서가 존재하는 이유
 
* Goal: Glusterfs에 대한 기술 정보 제공
* Scope: Glusterfs의 개념, 사용법, 설정
* Non-goals: 다른 주제로의 확장
 
 
== Key Concepts ==
 
{| class="wikitable"
! Concept
! Description
! Related
|-
| Glusterfs
| HPC/서버 환경에서 중요한 기술 개념
| [[Linux]], [[Server]]
|}
 
 
== Detailed Explanation ==
 
.glusterfs creates a hardlink for each of the file present in the brick. The contents of .glusterfs must not be touched unless you know very well what you are doing. Its apparent size is not real. It mostly contains hard-links to other files, so the net space usage contribution for this directory is very small.<ref>https://github.com/gluster/glusterfs/issues/833</ref>  
.glusterfs creates a hardlink for each of the file present in the brick. The contents of .glusterfs must not be touched unless you know very well what you are doing. Its apparent size is not real. It mostly contains hard-links to other files, so the net space usage contribution for this directory is very small.<ref>https://github.com/gluster/glusterfs/issues/833</ref>  
If you use 'du' on the entire volume, you will get the actual used space because 'du' already takes into account hard links.
If you use 'du' on the entire volume, you will get the actual used space because 'du' already takes into account hard links.
From the root of the brick, run this:<syntaxhighlight lang="bash">
From the root of the brick, run this:<syntaxhighlight lang="bash">
# du -sh * .glusterfs  
# du -sh * .glusterfs  
Line 9: Line 50:
# find .glusterfs -type f | wc -l
# find .glusterfs -type f | wc -l
</syntaxhighlight>  
</syntaxhighlight>  
[https://icicimov.github.io/blog/high-availability/GlusterFS-orphaned-GFID-hard-links/ Igor Cicimov] describes well how to get rid of GlusterFS orphaned GFID hard links <syntaxhighlight lang="bash">
[https://icicimov.github.io/blog/high-availability/GlusterFS-orphaned-GFID-hard-links/ Igor Cicimov] describes well how to get rid of GlusterFS orphaned GFID hard links <syntaxhighlight lang="bash">
# find /path-to-bric/.glusterfs -type f -links -2 -exec rm -fv {} \;
# find /path-to-bric/.glusterfs -type f -links -2 -exec rm -fv {} \;
</syntaxhighlight>
</syntaxhighlight>
== GlusterFS Performance Tuning ==
performance tuning, there are no magic values for these which work on all systems. The defaults in GlusterFS are configured at install time to provide best performance over mixed workloads. To squeeze performance out of GlusterFS, use an understanding of the below parameters and how them may be used in your setup.
performance tuning, there are no magic values for these which work on all systems. The defaults in GlusterFS are configured at install time to provide best performance over mixed workloads. To squeeze performance out of GlusterFS, use an understanding of the below parameters and how them may be used in your setup.
After making a change, be sure to restart all GlusterFS processes and begin benchmarking the new values.<ref>https://www.jamescoyle.net/how-to/559-glusterfs-performance-tuning</ref>
After making a change, be sure to restart all GlusterFS processes and begin benchmarking the new values.<ref>https://www.jamescoyle.net/how-to/559-glusterfs-performance-tuning</ref>
== General Commands ==
=== Check Status ===
<nowiki>#</nowiki> gluster peer status
<nowiki>#</nowiki> gluster peer status
<nowiki>#</nowiki> gluster --remote-host=nod1 peer status  
<nowiki>#</nowiki> gluster --remote-host=nod1 peer status  
<nowiki>#</nowiki> gluster pool list
<nowiki>#</nowiki> gluster pool list
=== Prove peer ===
<nowiki>#</nowiki> gluster peer probe <IP>
<nowiki>#</nowiki> gluster peer probe <IP>
=== Detach peer ===
<nowiki>#</nowiki> gluster peer detach <node>
<nowiki>#</nowiki> gluster peer detach <node>
<nowiki>#</nowiki> gluster peer detach <node> force
<nowiki>#</nowiki> gluster peer detach <node> force
=== Volume ===
<nowiki>#</nowiki> gluster volume create <Volume name> transport tcp <brick path>  
<nowiki>#</nowiki> gluster volume create <Volume name> transport tcp <brick path>  
<nowiki>#</nowiki> gluster volume create ucmvolume nod1:/blok1 nod2:/blok1
<nowiki>#</nowiki> gluster volume create ucmvolume nod1:/blok1 nod2:/blok1


<nowiki>#</nowiki> gluster volume list
# gluster volume get <Volume name> all


Option                                  Value                                 
== Best Practices ==
------                                  -----                                 
cluster.lookup-unhashed                on
...
<nowiki>#</nowiki> gluster volume '''reset''' <Volume name> nfs.disable


<nowiki>#</nowiki> gluster volume set ucmvolume nfs.disable on
* 최신 버전 사용 권장
* 공식 문서 참고
* 테스트 환경에서 먼저 검증


<nowiki>#</nowiki> gluster volume status <Volume name>


<nowiki>#</nowiki> gluster volume start <Volume name>
== References ==


<nowiki>#</nowiki> gluster volume info <Volume name>
* [https://wiki.hpcmate.com Glusterfs]


== Remove bricks ==
root@nod1 :~'''# gluster volume remove-brick ucmvolume nod1:/blok1 start'''
volume remove-brick start: success
ID: c6ab64f7-d921-4e07-9350-0524b7d2a613
root@nod1 :~# gluster volume remove-brick ucmvolume  nod1:/blok1 status
    Node  Rebalanced-files        size      scanned    failures      skipped      '''status'''  run time in secs
  ---------  -----------  -----------  -----------  -----------  -----------  ------------  --------------
    localhost            0      0Bytes            0            0            0    '''completed'''            0.00
root@nod1 :~'''# gluster volume remove-brick ucmvolume nod1:/blok1 commit'''
Removing brick(s) can result in data loss. Do you want to Continue? (y/n) y
volume remove-brick commit: success<br />
<nowiki>#</nowiki> mount -t glusterfs nod1:/ucmvolume /mnt


<nowiki>#</nowiki> mount -t glusterfs nod2:/ucmvolume /mnt -o backupvolfile-server=nod1
== Related Pages ==


<nowiki>#</nowiki> mount -t glusterfs nod2,nod1:/ucmvolume /mnt
* [[Linux]]
* [[Server]]
* [[Hardware]]
* [[Network]]


<nowiki>#</nowiki> gluster snapshot list# gluster snapshot status


<nowiki>#</nowiki> gluster snapshot info
[[Category:Linux]]
Snap Name : '''build1_GMT-2015.05.26-14.19.01'''
== Knowledge Graph ==
Snap UUID : 3b3b8e45-5c81-45ff-8e82-fea05b1a516a
        Brick Path       :  nod1:/run/gluster/snaps/25fa0c73a6b24327927c0dc9f4f08dba/brick1/data
        Volume Group     :  ThinGroup
        Brick Running     :  No
        Brick PID         :  N/A
        Data Percentage   :  40.80
        LV Size           :  80.00g
        Brick Path       :  nod2:/run/gluster/snaps/25fa0c73a6b24327927c0dc9f4f08dba/brick2/data
        Volume Group     :  ThinGroup
        Brick Running     :  No
        Brick PID         :  N/A
        Data Percentage   :  40.75
        LV Size           :  80.00g
        Brick Path       :  nod3:/run/gluster/snaps/25fa0c73a6b24327927c0dc9f4f08dba/brick3/data
        Volume Group     :  ThinGroup
        Brick Running     :  No
        Brick PID         :  N/A
        Data Percentage   :  40.82
        LV Size           :   80.00g
<nowiki>#</nowiki> gluster snapshot activate '''build1_GMT-2015.05.26-14.19.01'''


<nowiki>#</nowiki> gluster snapshot deactivate build1_GMT-2015.05.26-14.19.01
Related


<nowiki>#</nowiki> mount -t glusterfs localhost:/snaps/build1_GMT-2015.05.26-14.19.01/buildinglab /mnt
→ [[Linux]]
root@nod1~# gluster snapshot clone build-new build1_GMT-2015.05.26-14.19.01
→ [[Server]]
snapshot clone: success: Clone build-new created successfully
→ [[Hardware]]
root@nod1~# mount
→ [[Network]]
...
/dev/mapper/ThinGroup-build--new_0 on /srv/buildinglab type btrfs (rw,relatime,nodatasum,nodatacow,space_cache,autodefrag)
/dev/mapper/ThinGroup-build--new_0 on /run/gluster/snaps/25fa0c73a6b24327927c0dc9f4f08dba/brick2 type btrfs (rw,relatime,nodatasum,nodatacow,space_cache,autodefrag)
/dev/mapper/ThinGroup-build--new_0 on /run/gluster/snaps/build-new/brick2 type btrfs (rw,relatime,nodatasum,nodatacow,space_cache,autodefrag)
...


[[Category:Guide]]


<nowiki>#</nowiki> gluster volume heal <Volume name> enable
== GlusterFS Performance Tuning ==


<nowiki>#</nowiki> gluster volume heal git info
GlusterFS 성능 튜닝은 워크로드 특성에 따라 다름.


=== Read/Write Buffering ===


== GlusterFS Configuration ==
<syntaxhighlight lang="bash">
GlusterFS volumes can be configured with multiple settings. These can be set on a volume using the below command substituting '''[VOLUME]''' for the volume to alter, '''[OPTION]'''  for the parameter name and '''[PARAMETER]''' for the parameter value.
# /etc/glusterfs/glusterd.vol 수정
gluster volume set '''[VOLUME] [OPTION] [PARAMETER]'''
option rpc-auth.auth-unix.allow-missing-on # 성능 개선
Example:
option rpc-auth.auth-multipath on
gluster volume set '''myvolume performance.cache-size 1GB'''
option rpc-auth.allow-insecure on
Or you can add the parameter to the '''glusterfs.vol''' config file.
'''vi /etc/glusterfs/glusterfs.vol'''


* '''performance.write-behind-window-size''' – the size in bytes to use for the per file write behind buffer. Default: 1MB.
# Gluster volume 옵션
* '''performance.cache-refresh-timeout''' – the time in seconds a cached data file will be kept until data revalidation occurs. Default: 1 second.
gluster volume set <volname> performance.read-ahead on
* '''performance.cache-size''' – the size in bytes to use for the read cache. Default: 32MB.
gluster volume set <volname> performance.io-thread-count 32
* '''cluster.stripe-block-size''' – the size in bytes of the unit that will be read from or written to on the GlusterFS volume. Smaller values are better for smaller files and larger sizes for larger files. Default: 128KB.
gluster volume set <volname> network.ping-timeout 30
* '''performance.io-thread-count''' – is the maximum number of threads used for IO. Higher numbers improve concurrent IO operations, providing your disks can keep up. Default: 16.
</syntaxhighlight>


== Tunning Points<ref>https://docs.gluster.org/en/main/Administrator-Guide/Performance-Tuning/</ref> ==
=== Replication Optimization ===
When mounting your storage for the GlusterFS later, make sure it is configured for the type of [[workload]] you have.


=== Disable <code>performance.client-io-threads on distributed volumes</code> ===
<syntaxhighlight lang="bash">
Red Hat recommends disabling the <code>performance.client-io-threads</code> option on distributed volumes, as this option tends to worsen performance. Run the following<ref>https://access.redhat.com/documentation/ko-kr/red_hat_gluster_storage/3.4/html/administration_guide/creating_distributed_volumes</ref>
# 리플리케이션 스레드 수 증가
$ sudo gluster volume set ucmgv performance.client-io-threads off
gluster volume set <volname> performance.replica-io-thread-count 16


=== Enable Metadata cache ===
# 디스트리뷰션 스레드 수 증가
gluster volume set <volname> performance.dist-io-thread-count 32


# Enable metadata caching and cache invalidation: 
# RDMA 사용 (네트워크 대역폭↑)
## <code>gluster volume set <volname> group metadata-cache</code>  This group command enables caching of stat and xattr information of a file or directory. The caching is refreshed every 10 min, and cache-invalidation is enabled to ensure cache consistency.
gluster volume set <volname> server.rdma on
# To increase the number of files that can be cached, execute the following command: 
</syntaxhighlight>
## <code>gluster volume set <volname> network.inode-lru-limit <n></code>  n, is set to 50000. It can be increased if the number of active files in the volume is very high. Increasing this number increases the memory footprint of the brick processes.
# Enable [[samba]] specific metadata caching: 
## <code>gluster volume set <volname> cache-samba-metadata on</code>
# By default, some xattrs are cached by gluster like: capability xattrs, ima xattrs    ACLs, etc. If there are any other xattrs that are used by the application using    the Gluster storage, execute the following command to add these xattrs to the metadata    cache list:   
## <code>gluster volume set <volname> xattr-cache-list "comma separated xattr list"</code>   
## Eg:    <code>gluster volume set <volname> xattr-cache-list "user.org.netatalk.*,user.swift.metadata"</code>


=== Directory operations ===
=== Extended Attributes (xattr) ===
Along with enabling the metadata caching, the following options can be set to increase performance of directory operations:


==== Directory listing Performance: ====
<syntaxhighlight lang="bash">
 
# xattr 캐싱 활성화
* Enable <code>parallel-readdir</code>
gluster volume set <volname> network.inode-lru-limit 100000
 
gluster volume set <volname> server.allow-insecure on
<code>gluster volume set <VOLNAME> performance.readdir-ahead on  gluster volume set <VOLNAME> performance.parallel-readdir on</code>
</syntaxhighlight>
 
==== File/Directory Create Performance ====
 
* Enable <code>nl-cache</code>
 
<code>gluster volume set <volname> group nl-cache  gluster volume set <volname> nl-cache-positive-entry on</code>
 
The above command also enables cache invalidation and increases the timeout to 10 minutes
 
=== Small file Read operations ===
For use cases with dominant small file reads, enable the following options
gluster volume set <volname> performance.cache-invalidation on
gluster volume set <volname> features.cache-invalidation on
gluster volume set <volname> performance.qr-cache-timeout 600 # 10 min recommended setting
gluster volume set <volname> cache-invalidation-timeout 600 # 10 min recommended setting
This command enables caching of the content of small file, in the client cache. Enabling cache invalidation ensures cache consistency.
 
The total cache size can be set using
<code>gluster volume set <volname> cache-size <size></code>
By default, the files with size <code><=64KB</code> are cached. To change this value:
<code>gluster volume set <volname> performance.cache-max-file-size <size></code>
Note that the <code>size</code> arguments use SI unit suffixes, e.g. <code>64KB</code> or <code>2MB</code>.
 
=== Others ===
* When mounting your GlusterFS storage from a remote server to your local server, be sure to '''dissable''' direct-io '''as this will enable the kernel read ahead and file system cache'''. This will be sensible for most workloads where caching of files is beneficial.
* When mounting the GlusterFS volume over NFS use '''noatime''' and '''nodiratime''' to remove the timestamps over NFS.
 
* # gluster volume set $vol performance.o-thread-count 64<ref>https://www.spinics.net/lists/gluster-users/msg24680.html</ref>, Today’s CPU are powerful enough to handle 64 threads per volume.
 
* # gluster volume set $vol client.event-threads XX, XX depend on the number of connections from the FUSE client to the server, you can get this number by running netstat and grep on the server IP and count the number of connections.
 
* # gluster volume set $vol server.event-threads XX,  XX depend on the number of connections from the server to the client(s), you can get this number by running netstat and grep on “gluster" and count the number of connections.
 
== Troubleshooting ==
 
=== Glusterfs log location ===
/var/log/glusterfs
 
=== Peer Rejected (connected) state in gluster peer status ===
Hostname: <hostname>
Uuid: '''<xxxx-xxx-xxxx>'''
State: '''Peer Rejected (Connected)'''
 
==== Solution 1, Fully sync with the rest of the trusted cluster pool<ref>https://staged-gluster-docs.readthedocs.io/en/release3.7.0beta1/Administrator%20Guide/Resolving%20Peer%20Rejected/?q=remove+blick&check_keywords=yes&area=default</ref> ====
This indicates that the volume configuration on the node is not in sync with the rest of the trusted cluster storage pool.
 
# first check and compare the peers brick UUID # between master view and /var/lib/glusterd/glusterd.info on rejected peer. if the uuid is different then you need to update master's view uuid on step #2 in following procedure.
# On the rejected peer: (Try the whole procedure a couple more times if it doesn't work right away)
## Stop glusterd
## In /var/lib/glusterd, delete everything except glusterd.info (the    UUID file)
## Start glusterd
## Probe one of the good peers
## Restart glusterd, check 'gluster peer status'
## You may need to restart glusterd another time or two, keep checking    peer status


==== Solution 2, Update the cluster.op-version<ref>https://gitlab.ito.umt.edu/zr139734e/glusterdocs/-/blob/v3/Troubleshooting/troubleshooting-glusterd.md</ref> ====
=== 확장성 가이드 ===
 
* Run <code>gluster volume get all cluster.max-op-version</code> to get the latest supported op-version.
* Update the cluster.op-version to the latest supported op-version by executing <code>gluster volume set all cluster.op-version <op-version></code>.
 
==== Solution3, Expecting data loss ====
If you don't care about the data on the failed peer and just want to resync, the following might work (I take no responsibility for any data loss, etc, etc)<ref>https://github.com/gluster/glusterfs/issues/159</ref>
<# Remove dead server
gluster volume remove-brick myvolume 10.0.0.5:/brickpath force
gluster peer detach 10.0.0.5
# add peer again
gluster peer probe 10.0.0.5
gluster volume add-brick myvolume 10.0.0.5:/brickpath force
 
== Reference ==
[[Category:Linux]]


[[Category:Reference]]
{| class="wikitable"
! 구성 ! 최대 브릭 수 ! 최대 파일 수 ! 권장 용도
|-
| 4 브릭 디스트리뷰션 | 1,000 | 100M | 소규모 HPC
|-
| 8 브릭 리플리케이션 | 500 | 50M | 중규모 클러스터
|-
| 16 브릭 디스트리뷰션 | 2,000 | 500M | 대규모 HPC
|-
| 32+ 브릭 | 10,000+ | 1B+ | 엔터프라이즈
|}

Latest revision as of 13:37, 17 July 2026

Template:Status

Template:TOC

Overview

Glusterfs에 대한 기술 문서입니다.

Summary

  • 무엇인가? - Glusterfs
  • 왜 필요한가? - HPC 및 서버 환경에서 필수 개념
  • 언제 사용하는가? - 서버 구성, 성능 튜닝, 문제 해결 시


Purpose

이 문서가 존재하는 이유

  • Goal: Glusterfs에 대한 기술 정보 제공
  • Scope: Glusterfs의 개념, 사용법, 설정
  • Non-goals: 다른 주제로의 확장


Key Concepts

Concept Description Related
Glusterfs HPC/서버 환경에서 중요한 기술 개념 Linux, Server


Detailed Explanation

.glusterfs creates a hardlink for each of the file present in the brick. The contents of .glusterfs must not be touched unless you know very well what you are doing. Its apparent size is not real. It mostly contains hard-links to other files, so the net space usage contribution for this directory is very small.[1] If you use 'du' on the entire volume, you will get the actual used space because 'du' already takes into account hard links.

From the root of the brick, run this:

# du -sh * .glusterfs 
# find .glusterfs -type f -links 1      #check if there are any files that have just 1 link
# find .glusterfs -type f | wc -l

Igor Cicimov describes well how to get rid of GlusterFS orphaned GFID hard links 

# find /path-to-bric/.glusterfs -type f -links -2 -exec rm -fv {} \;

performance tuning, there are no magic values for these which work on all systems. The defaults in GlusterFS are configured at install time to provide best performance over mixed workloads. To squeeze performance out of GlusterFS, use an understanding of the below parameters and how them may be used in your setup. After making a change, be sure to restart all GlusterFS processes and begin benchmarking the new values.[2] # gluster peer status # gluster --remote-host=nod1 peer status # gluster pool list # gluster peer probe <IP> # gluster peer detach <node> # gluster peer detach <node> force # gluster volume create <Volume name> transport tcp <brick path> # gluster volume create ucmvolume nod1:/blok1 nod2:/blok1


Best Practices

  • 최신 버전 사용 권장
  • 공식 문서 참고
  • 테스트 환경에서 먼저 검증


References


Related Pages

Knowledge Graph

Related

→ Linux → Server → Hardware → Network

GlusterFS Performance Tuning

GlusterFS 성능 튜닝은 워크로드 특성에 따라 다름.

Read/Write Buffering

# /etc/glusterfs/glusterd.vol 수정
option rpc-auth.auth-unix.allow-missing-on # 성능 개선
option rpc-auth.auth-multipath on
option rpc-auth.allow-insecure on

# Gluster volume 옵션
gluster volume set <volname> performance.read-ahead on
gluster volume set <volname> performance.io-thread-count 32
gluster volume set <volname> network.ping-timeout 30

Replication Optimization

# 리플리케이션 스레드 수 증가
gluster volume set <volname> performance.replica-io-thread-count 16

# 디스트리뷰션 스레드 수 증가
gluster volume set <volname> performance.dist-io-thread-count 32

# RDMA 사용 (네트워크 대역폭↑)
gluster volume set <volname> server.rdma on

Extended Attributes (xattr)

# xattr 캐싱 활성화
gluster volume set <volname> network.inode-lru-limit 100000
gluster volume set <volname> server.allow-insecure on

확장성 가이드

구성 ! 최대 브릭 수 ! 최대 파일 수 ! 권장 용도
1,000 | 100M | 소규모 HPC
500 | 50M | 중규모 클러스터
2,000 | 500M | 대규모 HPC
10,000+ | 1B+ | 엔터프라이즈
  1. ↑ https://github.com/gluster/glusterfs/issues/833
  2. ↑ https://www.jamescoyle.net/how-to/559-glusterfs-performance-tuning
Retrieved from "http://wiki.hpcmate.com/index.php?title=Glusterfs&oldid=3522"