Ogy
commented
6 years ago Hello, thanks for suggestion. I'm using it quite differently. I have mounted remote filesystem on one node, where is in place cron, so I will everytime use cron on one node. Due to RDB it doesn't matter where VM runs. What the --only-local should do ? Is there any reason to have synchronized cron over all nodes (i can imagine one - the node failure) What would be really interesting would be to use hook script to check or mount that remote FS (OK, i've just read docs,and there is --script )
I have two questions for you... 1) Today, whole Data Center went down, for awhile, before cron started backup. so remote FS has been disconnected, then BARC started due to low space on local "/" the backup failed the question is: will next backups be "corrupted" ? (i have --keep 31(it will miss one backup for recovery(am I correct?))) 2) can you, please, modify the lock setting? or introduce something like backup queue snapshot function has been modified to do snapshots of all disks of a VM at once (again thanks for it !). so if i'm correct, there is no needs to lock (let fail) other sheduled backups (other VMs) yeah, we need to let backup fail - when its exports take more than 23:59 hours i have few TB VMs, so one full backup is blocking backups of small VMs. I can recommend to use ZFS as storage for backups, because it doesn't store zeroes Unfortunately when I'm doing recovery, the !@#$ zeroes are being transfered over network, so it takes a lot of time.
Many thanks for your time and great BARC Thanks Tomas
Dne 12.10.2017 v 17:50 Daniele Corsini napsal(a):
Hello to all,
Now the cron file is in /etc/cron.d. My idea is to move it to /etc/pce/eve/barc/barc.cron so that it is replicated across the entire cluster (corosync). In this way you can have a unique configuration and with the --only-local parameter the node will only backup its vm, naturally following the rules of the --vmid. If vm is moved to another node, there will always be backup without having to reconfigure.
What do you think?
Thank you all
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/EnterpriseVE/eve4pve-barc/issues/18, or mute the thread https://github.com/notifications/unsubscribe-auth/ABFEabpaEIuMiAK1GdjytTIyrF1391JGks5srjU4gaJpZM4P3OgB.
franklupo
gardar
Hello to all,
Now the cron file is in /etc/cron.d. My idea is to move it to /etc/pce/eve/barc/barc.cron so that it is replicated across the entire cluster (corosync). In this way you can have a unique configuration and with the --only-local parameter the node will only backup its vm, naturally following the rules of the --vmid. If vm is moved to another node, there will always be backup without having to reconfigure.
What do you think?
Thank you all