Recent content by nes

Hi, had to check some other things first ... SSD-TBW etc. - yeah i am trying to get some (lightly) used 400GB S3700's so even more TBW to burn :-) i plan to use one HD for the Proxmox root disk *and* for the Ceph OSD - if posible (and not a dumb idea) if not possible (or dumb) - i will buy a...

Hi Q-wulf Hi udo thanks for the replies Will have more questions tomorrow :-)

IB : https://pve.proxmox.com/wiki/Infiniband doesnt sound too complicated at first sight .. but the 40GbE cards i wanted are gone - will have to see if i want to import it from the UK or US (?). QoS : So if i have 2 switches and one fails then the clients connected to the second switch will...

SSD : ok, will then buy the S3700 200GB (3PB TBW) - costs more money than i really have, but better this than a broken installation ... (and if i leave 10-20% empty then the controller can use this to extend the TBW even more, right ?) i read up on the threads about SSD failing with PVE...

So my revised plan is : 3 Hosts with CPU : i5 3450 (one i5 2500) RAM : 32GB HDD : WD Re 2TB or something similar (WD Gold ?) no ZFS to save some RAM (as Ceph should be good enough) SSD : Intel SSD DC S3700 200GB (3PB TBW) or Intel SSD DC S3610 200GB (1,1 PB TBW) still unsure about the...

Hi, €/TBW Ok, thanks. TBW is usually not a hard limit on SSDs (as far as i gathered from different tests) but of course it feels safer to have some guaranteed numbers. So i will need to spend quite a bit more on the SSDs ... switch - well that gets even more expensive . The lowest cost...

Hi Udo, thanks for the response. 32GB - sadly no way to expand this. (except buying new Servers) i could run the vms spread out on the 3 hosts and only if one fails then it would get cramped on one of the remaining hosts. Is it possible to define a failover with less memory for the vm than it...

Hallo, i am planning to build a 3 node proxmox ceph cluster and am looking for advice if my choices are good enough for a (very) small corporate cluster. Basically it would run a Windows-AD Server and a Linux-Postgres LXC (both serving ~50 Clients) another LXC with Nextcloud (low traffic)...