Re: [Gluster-users] Bitrot strange behavior
Hi Cedric, Any file is picked up for signing by the bitd process after the predetermined wait of 120 seconds. This default value is captured in the volume option 'features.expiry-time' and is configurable - in your case, it can be set to 0 or 1. Point 2 is correct. A file corrupted before the bitrot signature is generated will not be successfully detected by the scrubber. That would require admin/manual intervention to explicitly heal the corrupted file. -Sweta On 04/16/2018 10:42 PM, Cedric Lemarchand wrote: Hello, I am playing around with the bitrot feature and have some questions: 1. when a file is created, the "trusted.bit-rot.signature” attribute seems only created approximatively 120 seconds after its creations (the cluster is idle and there is only one file living on it). Why ? Is there a way to make this attribute generated at the same time of the file creation ? 2. corrupting a file (adding a 0 locally on a brick) before the creation of the "trusted.bit-rot.signature” do not provide any warning: its signature is different than the 2 others copies on other bricks. Starting a scrub did not show up anything. I would think that Gluster compares signature between bricks for this particular use cases, but it seems the check is only local, so a file corrupted before it’s bitrot signature creation stay corrupted, and thus could be served to clients whith bad data ? Gluster 3.12.8 on Debian Stretch, bricks on ext4. Volume Name: vol1 Type: Replicate Volume ID: 85ccfaf2-5793-46f2-bd20-3f823b0a2232 Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: gluster-01:/data/brick1 Brick2: gluster-02:/data/brick2 Brick3: gluster-03:/data/brick3 Options Reconfigured: storage.build-pgfid: on performance.client-io-threads: off nfs.disable: on transport.address-family: inet features.bitrot: on features.scrub: Active features.scrub-throttle: aggressive features.scrub-freq: hourly Cheers, Cédric ___ Gluster-users mailing list Gluster-users@gluster.org http://lists.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://lists.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] bit-rot resolution
Bitrot has a scrub process which detects the corrupted file. Once detected, it is the prerogative of the user to follow the sequence of steps [1] to trigger a heal. Having said that, client access to that file is not impacted as it continues to serve data from the good copy. Steps remain the same irrespective of a replica 2 or replica 3 volume. [1] https://access.redhat.com/documentation/en-US/Red_Hat_Storage/3.1/html/Administration_Guide/ch20s03.html Thanks, Sweta On 02/28/2017 06:28 PM, Gandalf Corvotempesta wrote: In a replica 3, what happens in case of bit-rot detection on a file ? Is gluster smart enough to detect this and automatically heal the corrupted files from other replicas ? What if in case of replica 2 ? How do know know which is right, server1 or server2, without a quorum ? What if the underling FS (like ZFS) is retuning an error in case of bit-rot ? ZFS should return an error if file is corrupted (and has not RAID to recover from), thus gluster should see the file as missing/corrupted and automatically trigger selfheal from other replicas ? ___ Gluster-users mailing list Gluster-users@gluster.org http://lists.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://lists.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] How to raed / write in geo-replication slave
Hi Milos, If I get this correctly, this is your configuration below: * Geo-rep master- Europe * Geo-rep slave - Asia * Same web application installed, and users from Asia are redirected to slave (in Asia), users from rest of the world are directed to master (in Europe) And you want read/write facility to be available in slave as well. YES, this is possible. BUT, it defeats the purpose of geo-replication. As per the concept of disaster recovery, read/writes should be carried out only at the master, and slave (if at all used), should be limited to only read workload. In your case, the very first write to the slave is going to render your geo-rep setup out-of-sync, and you would eventually have to delete off all the (new) content at the slave to get back the master and slave in sync. 1. If you are okay with this and would still like to go ahead with writes at both master and slave, all you would have to do is get the slave volume online. 2. Alternative: IF you are okay with writes happening at only one of the place at a time, then you could alternate between making Asia as master(and Europe as slave) and Asia as slave (and Europe as master), with the steps mentioned in this section of the Admin guide: http://documentation-devel.engineering.redhat.com/site/documentation/en-US/Red_Hat_Storage/3/html-single/Administration_Guide/index.html#Promoting_a_Slave_to_Master Would like to reiterate again - (2) is fine.. (1) is NOT recommended. Do reach out for any other queries. Thanks, Shweta On 07/14/2015 07:57 PM, Milos Cuculovic - MDPI wrote: Hi All, I have two GlusterFS servers, one master and one slave. They are doing geo-replication. I need to read and write on both of them. Here are some details: 2 data centers (One in Europe and one in Asia) The same web application is installed on both data centers Users from Asia are redirected to the Asian DC Users from the rest of the World and redirected in Europe The application is r/w some files from the glusterFs storage server The master is in Europe, and the r/w is fast I need people from the Asian DC to also r/w directly in Asia Is this possible, if so, any help please? I appriciate your help. Milos ___ 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 mailing list Gluster-users@gluster.org http://www.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] How to raed / write in geo-replication slave
Hi Milos, If I get this correctly, this is your configuration below: * Geo-rep master- Europe * Geo-rep slave - Asia * Same web application installed, and users from Asia are redirected to slave (in Asia), users from rest of the world are directed to master (in Europe) And you want read/write facility to be available in slave as well. YES, this is possible. BUT, it defeats the purpose of geo-replication. As per the concept of disaster recovery, read/writes should be carried out only at the master, and slave (if at all used), should be limited to only read workload. In your case, the very first write to the slave is going to render your geo-rep setup out-of-sync, and you would eventually have to delete off all the (new) content at the slave to get back the master and slave in sync. 1. If you are okay with this and would still like to go ahead with writes at both master and slave, all you would have to do is get the slave volume online. 2. Alternative: IF you are okay with writes happening at only one of the place at a time, then you could alternate between making Asia as master(and Europe as slave) and Asia as slave (and Europe as master), with the steps mentioned in this section of the Admin guide: http://documentation-devel.engineering.redhat.com/site/documentation/en-US/Red_Hat_Storage/3/html-single/Administration_Guide/index.html#Promoting_a_Slave_to_Master Would like to reiterate again - (2) is fine.. (1) is NOT recommended. Do reach out for any other queries. Thanks, Shweta On 07/14/2015 07:57 PM, Milos Cuculovic - MDPI wrote: Hi All, I have two GlusterFS servers, one master and one slave. They are doing geo-replication. I need to read and write on both of them. Here are some details: 2 data centers (One in Europe and one in Asia) The same web application is installed on both data centers Users from Asia are redirected to the Asian DC Users from the rest of the World and redirected in Europe The application is r/w some files from the glusterFs storage server The master is in Europe, and the r/w is fast I need people from the Asian DC to also r/w directly in Asia Is this possible, if so, any help please? I appriciate your help. Milos ___ 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
Re: [Gluster-users] Gluster peer status disconnected
Hi Shyam, Were the peers in a connected state anytime *before*? If yes, then these are the things you can do to troubleshoot and get the cluster back to normal (some of which Neils has already mentioned ): 1. Check if the nodes are reachable.. 'ping ipaddress' 2. Check if the glusterd daemon is in an active state in all the nodes.. 'service glusterd status' 3. Firewall issue - flush IPtables on the node which is disconnected.. 'iptables -F' Do let us know if this helps.. Regarding *how* you landed up in this situation, as Atin mentioned, it would be better to be a little specific on the things that you did at your end. Thanks, Shweta On 04/19/2015 06:45 PM, Niels de Vos wrote: On Sun, Apr 19, 2015 at 03:04:15PM +0530, Shyam Deshmukh wrote: Hi, I am getting following status and my cluster is down due to the same. Please help. gluster@gluster1:~$ sudo gluster peer status Number of Peers: 3 Hostname: gluster3 Uuid: c6f6574b-9779-4635-a7c2-06185a9ae973 State: Peer in Cluster (Disconnected) Hostname: gluster4 Uuid: c0e2ad61-da65-4e4c-a863-e9e8c4e84710 State: Peer in Cluster (Disconnected) Hostname: gluster2 Uuid: 45ac276d-9e65-4e32-8abe-51f8ba0241ce State: Peer in Cluster (Disconnected) gluster@gluster1:~$ You can probably start by verifying that the hostnames exactly as listed above can get resolved to their correct IP-addresses. If that works fine, check if you can connect to port 24007 on those systems. This might be a network issue, firewall or glusterd is not running (anymore) on those systems. Sometimes it also helps to check on other systems, and se if the output of gluster peer status is different there. Good luck! ___ 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