Search results for query: PLP

...and the rest of the space can be used for cold storage). It's the (many) VMs that need fast drives with lots of IOPS, like enterprise SSDs with PLP. RAM is (almost) always the limiting factor, as Proxmox does NOT work well with memory overcommit, and 32GiB is not much nowadays. (EDIT: NOT...

The Micron 7450 MAX are tlc with plp which is good. With "ZFSPOOL RAIDZ2" you get in fio just the iops of a single disk, recreate pool as mirror. For ceph you just have a single 25Gb active line for writing default 3 times each block so it's only 8Gbit effective. For linbit don't know your config.

...won't be written to the special devices, only newly written metadata will be stored on the special devices. If you decide to add special devices to your hdd pool then make sure they are good enterprise ssd's with PLP and at least 1DWPD, if you loose your special devices then all data is gone ...

RI is enough? I’m more expert on nutanix and they require only MU drives. In your opinion with 2 drives as cache how many performance can I get? It’s normal to get less than 1 normal HDD performance ?

...down. Don't use consumer grade stuff with zfs if you can. Depending on your workload you don't need mixed use ssd's, read intensive (but with PLP) could be enough. Would not use btrfs, while it's fine for single drive desktop workloads, raid seems to be unstable and recovering from that a...

...In a homelab with lvm thin / ext4 you should not have to worry too much about TBW but if you have to budget you could always get some (maybe used?) enterprise ssd's with PLP. They often have more TBW and better caching (with help of PLP) to migitate the wearout better. Maybe this reddit post...

Try removing the spinning drives. And per other threads PLP can make a big difference. What are the network speeds? 10 Gbps at least? Re crush, a given block/PG is copied to one drive on each host. You can restrict by drive type or set a primary read, or have a failure domain of drive instead...

I have a 3 node cluster. Has a bunch of drives, 1TB cold rust, 512 warm SataSSD And three 512 non-PLP NVMe that are Gen3. (1 Samsung SN730 and 2 Inland TN320) I know not to expect much - this is pre-prod - plan is to eventually get PLPs next year. 10Gb Emulex CNA is working very well with FRR...

...use case demands high continuous iops, you will want a different solution. if its for homelab they work just fine. 3. consumer drives do not have PLP (power loss protection.) what that means is that they're susceptible to data loss on power failure, but understand the implications. the "data...

@cwt Thanks for taking the time to break it all down in such detail, I really appreciate it. Thanks, I'll have a read while I wait for my Enterprise drives to arrive.

...once the SLC cache is gone. P4610 delivers a steady ~3,000 MB/s 24/7 without falling off. Power-loss protection: 990 Pro = none. P4610 = full PLP with capacitors, so in-flight data isn’t lost. Latency consistency: 990 Pro can spike to 10–50 ms under load. P4610 stays in the low 100 µs range...

more than one OSE per nvme will not help. non PLP are really like 500iops 4k write vs 20000iops for plp drive. At minimum, use cache=writeback, it should help to avoid small write when possible (merging small adjacent writes to big write)

I have a 3 node and 3 nvme gen3 (each) - non PLP ( I know - this is pre-prod ) My plan is to eventually get PLPs and until I can just backup daily to local storage. For the 3/2 RBD each nvme has one OSD. Performance is liek 10-20MBps writes - Win10 AsyncIO=threads,cache=non,discard/iothread=on...

...tests (power loss, forced resets, cable disconnects) I observed that when Ceph OSDs were located on SSDs without power-loss protection (PLP), virtual machines would occasionally become unbootable due to storage corruption. After migrating the OSDs to enterprise-grade SSDs with PLP (even...

The theory was double redundancy for hardware failure but the IO performance is such a punish I would be better off moving to 2 separate mirrors. Noted, I started with code then second guessed myself and changed it to quotes. Back to the drawing board to figure out the best way forward....

...for QLC on the forum and read about all the people who have a hard time believing they wasted their money (and tips about enterprise drives with PLP). Adding SLOG and/or L2ARC usually does not help but a special device might sometimes, but probably not in your case with already all SSDs...

Dont use consumer SSD without plp with zfs.

I would try again with "Enterprise Class SSDs" with PLP / "Power-Loss-Protection". ...and with mirrors, not a RaidZ2 - as this gives you only the IOPS of a single device. Of course SATA is massively slower than PCIe --> if possibly switch technology... Disclaimer: just random thoughts...

...(1G, corosync) and even more for the vm uplink (LACP / MLAG). A 3 node ceph cluster is for home lab / evaluation, not really someting you would want for production. The SSD's are certainly a good investment, consumer grade drives have some issues with ceph / zfs (fast wearout, no PLP, ...)

Relevante Details sind power-loss-protection ( plp ) und DWPD, da kann man bei Geizhals nach filtern und mal schauen, ob da die von dir ausgewählten Samsungmodelle auftauchen: https://geizhals.de/?cat=hdssd&xf=7161_Power-Loss Protection&promode=true Warum man das haben will, wurde mal...