Den ons 14 sep. 2022 kl 08:54 skrev gagan tiwari <gagan.tiwari@xxxxxxxxxxxxxxxxxx>: > Hi Guys, > I am new to Ceph and storage. We have a requirement of > managing around 40T of data which will be accessed by around 100 clients > all running RockyLinux9. > > We have a HP storage server with 12 SDD of 5T each and have set-up hardware > RAID6 on these disks. You have only one single machine? If so, run zfs on it and export storage as NFS. > HP storage server has 64G RAM and 18 cores. > > So, please advise how I should go about setting up Ceph on it to have best > read performance. We need fastest read performance. With NFSv4.x you can have local caching in the NFS client, that might help a lot for read perf if those 100 clients have local drives also. The reason I am not advocating ceph in this case is that ceph is built to have many servers feed data to many clients (or many processes doing separate reads) and you seem to have a "single-server" setup and in this case, the overhead of the ceph protocol will lower the performance compared to "simpler" solutions like NFS which are not designed to scale in the way ceph is. A smaller point is that for both zfs and ceph, it is not advisable to first raid the separate drives and then present them to the filesystem/network, but rather give zfs/ceph each individual disk to handle it at a higher level. But compared to the "I have one or I have many servers to serve file IO" it is a small thing. -- May the most significant bit of your life be positive. _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |