This keeps coming up, which is not surprising, considering it is a core question. Here's how I look at it: The Ceph team has chosen to default to N+2 redundancy. This is analogous to RAID 6 (NOT RAID 1). The basic reasoning for N+2 in storage is as follows: If you experience downtime (either routine, or non-routine), then you can survive an additional failure during recovery. The scenario goes like this: Update software on OSD host Reboot All hosted OSDs are marked down Host come online, hosted OSDs come online Recovery begins Drive in another host is found to have an unreadable sector If you only have single redundancy (N+1, i.e. R2, m=X n=1, RAID1, RAID5), then you have now lost data. If you have double redundancy (N+2, R3, m=X n=2, RAID6), then there is a third method to get the good data, and both redundancy layers can be rebuilt. RAID6 came into being because the longer you spend recovering, the more likely you are to run into an undetected failure. Ceph takes the same view; it is built for MASSIVE storage, with LONG recovery times. My pools are built on 10Tb drives, others in this list use 12Tb and 14Tb drives. It all comes down to this: are you absolutely certain that you will never encounter a latent failure, while executing routine maintenance? If you look at it from the stand point of Risk Management, using Dr. Reason's Swiss Cheese model, then each redundancy layer is a barrier against failure. Thank you, Dominic L. Hilsbos, MBA Director - Information Technology Perform Air International, Inc. DHilsbos@xxxxxxxxxxxxxx www.PerformAir.com -----Original Message----- From: Zhenshi Zhou [mailto:deaderzzs@xxxxxxxxx] Sent: Thursday, July 9, 2020 7:11 PM To: ceph-users Subject: about replica size Hi, As we all know, the default replica setting of 'size' is 3 which means there are 3 copies of an object. What is the disadvantages if I set it to 2, except I get fewer copies? Thanks _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |