[Gluster-users] Arbiter Addition in Replicated environment
Hi Team, Can we add Arbiter brick in 2 node replication running environment. For an example Glusterfs 2 node replication Current glusterfs storage size 4 TB. After adding Arbiter brick in this environment what will be the result. #gluster volume add-brick test replica 3 arbiter 1 server3:/glusterfs/arbi Thank You Atul Yadav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] Glusterfs readonly Issue
Dear Team, In the event of the failure of master1, master 2 glusterfs home directory will become read only fs. If we manually shutdown the master 2, then there is no impact on the file system and all io operation will complete with out any problem. can you please provide some guidance to isolate the problem. # gluster peer status Number of Peers: 2 Hostname: master1-ib.dbt.au Uuid: a5608d66-a3c6-450e-a239-108668083ff2 State: Peer in Cluster (Connected) Hostname: compute01-ib.dbt.au Uuid: d2c47fc2-f673-4790-b368-d214a58c59f4 State: Peer in Cluster (Connected) # gluster vol info home Volume Name: home Type: Replicate Volume ID: 2403ddf9-c2e0-4930-bc94-734772ef099f Status: Started Number of Bricks: 1 x 2 = 2 Transport-type: tcp,rdma Bricks: Brick1: master1-ib.dbt.au:/glusterfs/home/brick1 Brick2: master2-ib.dbt.au:/glusterfs/home/brick2 Options Reconfigured: performance.quick-read: off performance.read-ahead: off performance.io-cache: off performance.stat-prefetch: off network.remote-dio: enable cluster.quorum-type: auto nfs.disable: on performance.readdir-ahead: on cluster.server-quorum-type: server config.transport: tcp,rdma network.ping-timeout: 10 cluster.server-quorum-ratio: 51% cluster.enable-shared-storage: disable # gluster vol heal home info Brick master1-ib.dbt.au:/glusterfs/home/brick1 Status: Connected Number of entries: 0 Brick master2-ib.dbt.au:/glusterfs/home/brick2 Status: Connected Number of entries: 0 # gluster vol heal home info heal-failed Gathering list of heal failed entries on volume home has been unsuccessful on bricks that are down. Please check if all brick processes are running[root@master2 Thank You Atul Yadav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Gluster Peer behavior
Hi , After restarting the service. service entered in to the fail state. [root@master1 ~]# /etc/init.d/glusterd restart Stopping glusterd: [FAILED] Starting glusterd: [FAILED] Note: This behavior only happening over rdma network. But with ethernet there is no issue. Thank you Atul Yadav On Tue, Jul 5, 2016 at 11:28 AM, Atin Mukherjee wrote: > > > On Tue, Jul 5, 2016 at 11:01 AM, Atul Yadav > wrote: > >> Hi All, >> >> The glusterfs environment details are given below:- >> >> [root@master1 ~]# cat /etc/redhat-release >> CentOS release 6.7 (Final) >> [root@master1 ~]# uname -r >> 2.6.32-642.1.1.el6.x86_64 >> [root@master1 ~]# rpm -qa | grep -i gluster >> glusterfs-rdma-3.8rc2-1.el6.x86_64 >> glusterfs-api-3.8rc2-1.el6.x86_64 >> glusterfs-3.8rc2-1.el6.x86_64 >> glusterfs-cli-3.8rc2-1.el6.x86_64 >> glusterfs-client-xlators-3.8rc2-1.el6.x86_64 >> glusterfs-server-3.8rc2-1.el6.x86_64 >> glusterfs-fuse-3.8rc2-1.el6.x86_64 >> glusterfs-libs-3.8rc2-1.el6.x86_64 >> [root@master1 ~]# >> >> Volume Name: home >> Type: Replicate >> Volume ID: 2403ddf9-c2e0-4930-bc94-734772ef099f >> Status: Stopped >> Number of Bricks: 1 x 2 = 2 >> Transport-type: rdma >> Bricks: >> Brick1: master1-ib.dbt.au:/glusterfs/home/brick1 >> Brick2: master2-ib.dbt.au:/glusterfs/home/brick2 >> Options Reconfigured: >> network.ping-timeout: 20 >> nfs.disable: on >> performance.readdir-ahead: on >> transport.address-family: inet >> config.transport: rdma >> cluster.server-quorum-type: server >> cluster.quorum-type: fixed >> cluster.quorum-count: 1 >> locks.mandatory-locking: off >> cluster.enable-shared-storage: disable >> cluster.server-quorum-ratio: 51% >> >> When my single master node is up only, but other nodes are still showing >> connected mode >> gluster pool list >> UUIDHostnameState >> 89ccd72e-cb99-4b52-a2c0-388c99e5c7b3master2-ib.dbt.au Connected >> d2c47fc2-f673-4790-b368-d214a58c59f4compute01-ib.dbt.au Connected >> a5608d66-a3c6-450e-a239-108668083ff2localhost Connected >> [root@master1 ~]# >> >> >> Please advise us >> Is this normal behavior Or This is issue. >> > > First of, we don't have any master slave configuration mode for gluster > trusted storage pool i.e. peer list. Secondly, if master2 and compute01 are > still reflecting as 'connected' even though they are down it means that > localhost here didn't receive disconnect events for some reason. Could you > restart glusterd service on this node and check the output of gluster pool > list again? > > > >> >> Thank You >> Atul Yadav >> >> >> ___ >> Gluster-users mailing list >> Gluster-users@gluster.org >> http://www.gluster.org/mailman/listinfo/gluster-users >> > > ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] Gluster Peer behavior
Hi All, The glusterfs environment details are given below:- [root@master1 ~]# cat /etc/redhat-release CentOS release 6.7 (Final) [root@master1 ~]# uname -r 2.6.32-642.1.1.el6.x86_64 [root@master1 ~]# rpm -qa | grep -i gluster glusterfs-rdma-3.8rc2-1.el6.x86_64 glusterfs-api-3.8rc2-1.el6.x86_64 glusterfs-3.8rc2-1.el6.x86_64 glusterfs-cli-3.8rc2-1.el6.x86_64 glusterfs-client-xlators-3.8rc2-1.el6.x86_64 glusterfs-server-3.8rc2-1.el6.x86_64 glusterfs-fuse-3.8rc2-1.el6.x86_64 glusterfs-libs-3.8rc2-1.el6.x86_64 [root@master1 ~]# Volume Name: home Type: Replicate Volume ID: 2403ddf9-c2e0-4930-bc94-734772ef099f Status: Stopped Number of Bricks: 1 x 2 = 2 Transport-type: rdma Bricks: Brick1: master1-ib.dbt.au:/glusterfs/home/brick1 Brick2: master2-ib.dbt.au:/glusterfs/home/brick2 Options Reconfigured: network.ping-timeout: 20 nfs.disable: on performance.readdir-ahead: on transport.address-family: inet config.transport: rdma cluster.server-quorum-type: server cluster.quorum-type: fixed cluster.quorum-count: 1 locks.mandatory-locking: off cluster.enable-shared-storage: disable cluster.server-quorum-ratio: 51% When my single master node is up only, but other nodes are still showing connected mode gluster pool list UUIDHostnameState 89ccd72e-cb99-4b52-a2c0-388c99e5c7b3master2-ib.dbt.au Connected d2c47fc2-f673-4790-b368-d214a58c59f4compute01-ib.dbt.au Connected a5608d66-a3c6-450e-a239-108668083ff2localhost Connected [root@master1 ~]# Please advise us Is this normal behavior Or This is issue. Thank You Atul Yadav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] HPC with Glusterfs
Hi All, We are using glusterfs in our HPC environment. Infra details. Master Node 2 Compute Node 10 Storage: glusterfs3.8rc2 Operating System: CentOS 6.7 Infiniband: RDMA For achieving the high availability between both the master nodes glusterfs-server is setup in replicated mode. And all 10 compute nodes glusterfs client. Gluster peer created over IPoIB. HOME directory on glusterfs replicated over RDMA. Please share the fine tuning for the environment. Thank You Atul Yadav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] GlusterFS RDMA Support
Hi Team, We installed glusterfs3.8 in our HPC environment. While configuring RDMA on glusterfs, the erroe is coming. Is glusterfs rdma is compatible with OFED, Intel and Mellanox driver. Or It is only compatible with operating system infiniband driver. Thank You Atul YAdav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] GlusterFS First Time
Thanks for the reply. As per your guidance, the glusterfs information is given below: Volume Name: share Type: Replicate Volume ID: bd545058-0fd9-40d2-828b-7e60a4bae53c Status: Started Number of Bricks: 1 x 2 = 2 Transport-type: tcp Bricks: Brick1: master1:/data/brick1/share Brick2: master2:/data/brick2/share Options Reconfigured: performance.readdir-ahead: on cluster.self-heal-daemon: enable network.ping-timeout: 25 cluster.eager-lock: on [root@master1 ~]# gluster pool list UUIDHostnameState 5a479842-8ee9-4160-b8c6-0802d633b80fmaster2 Connected 5bbfaa4a-e7c5-46dd-9f5b-0a44f1a583e8localhost Connected Host information is given below:- 192.168.10.103 master1.local master1 #Fixed 192.168.10.104 master2.local master2 #Fixed Test case is given below:- Case 1 While writing 20MB of files continuously on client side. one of the glusterfs server(Master1) is powered off. Impact Client IO operation will be on hold for 25 to 30 second and after that it will work normally. Case 2 When failed server is power-up during the IO operation at client side. Impact Client IO operation will be on hold for 25 to 30 second and after that it will work normally. Result: There will be no IO loss during the event of failure. But there is difference of data size on both the servers. *Master1* *Size */dev/mapper/brick1-brick1 19912704 508320 19404384 3% /data/brick1 *Inodes */dev/mapper/brick1-brick19961472 1556 99599161% /data/brick1 *Master2* *Size* /dev/mapper/brick2-brick2 19912704 522608 19390096 3% /data/brick2 *Inodes */dev/mapper/brick2-brick29961472 1556 99599161% /data/brick2 *Client* *Size *master1.local:/share 19912704 522624 19390080 3% /media *Inodes *master1.local:/share 9961472 1556 99599161% /media How we can match the size of data on both the servers or this normal behavior. And there will be impact on data integrity. Thank You Atul Yadav 09980066464 On Fri, Feb 12, 2016 at 1:21 AM, Gmail wrote: > find my answers inline. > > *— Bishoy* > > On Feb 11, 2016, at 11:42 AM, Atul Yadav wrote: > > HI Team, > > > I am totally new in Glusterfs and evaluating glusterfs for my requirement. > > Need your valuable input on achieving below requirement :- > File locking > > Gluster uses DLM for locking. > > Performance > > It depends on your work load (is it small files big files, etc…), the > number of drives and the kind of volume you create. > I suggest you start with just a Distributed Replicated volume and from > that point you can plan for the hardware and software configuration. > > High Availability > > I suggest replicating the bricks across the two nodes, as erasure coding > with two nodes and a single drive on each one will not be of any benefit. > > > > Existing Infra details is given below:- > CentOS 6.6 > glusterfs-server-3.7.8-1.el6.x86_64 > glusterfs-client-xlators-3.7.8-1.el6.x86_64 > GlusterFS server Quantity 2 with independent 6 TB storage > 24 Glusterfs Client. > Brick replication > > > Thank You > Atul Yadav > > > ___ > Gluster-users mailing list > Gluster-users@gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users > > > ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] GlusterFS First Time
HI Team, I am totally new in Glusterfs and evaluating glusterfs for my requirement. Need your valuable input on achieving below requirement :- File locking Performance High Availability Existing Infra details is given below:- CentOS 6.6 glusterfs-server-3.7.8-1.el6.x86_64 glusterfs-client-xlators-3.7.8-1.el6.x86_64 GlusterFS server Quantity 2 with independent 6 TB storage 24 Glusterfs Client. Brick replication Thank You Atul Yadav ___ Gluster-users mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users