Search results

You should keep the mounter count to an un-even number. CEPH suggest only 3 unless you have a huge enviornment. So your current amount of MON's is much higher than the recommend

The issue is PVE-firewall is not able to start / be configured causing other services to fail. Is there anything in the logs directory or the status output of pve-firewall?

If your using LXC then the load shown in the VM will always be the load value on the node itself. There is talks of working on a feature to show a VM load value but nothing yet has been made of this.

Which virt are you using?

After

Journal only needs to be 5-10GB, however you need a decent SSD for the job, something like : http://www.samsung.com/us/business/computing/solid-state-drives/MZ-7KM240E

If you can get 6 OSD's of the same type and then one or two SSD's per a server for the Journal then your be able to improve performance, with the current hardware you have and without any changes not too much you can do.

In your current setup (amount of OSD's) the most you will get the is the max rating of a single OSD disk, however you will never hit 100% due to overheads and different performance ratings at different queue depths and I/O type.

A standard 7.5K disk will give you around 150, the 505 result you got will be due to a Raid Cache or others means.

In the current version of CEPH the underlying storage is a system called filestore, this means that every write is also wrote to a journal, this causes a thing called a double write penalty, basically for every write if it wrote twice, causing filestore to max out at 1/2 the speed of the...

A cluster needs 3 servers, so your need a small 3rd server with Proxmox installed.

As previously mentioned Proxmox is not really designed for that use case so is not something you can do via the GUI. However you technically can pass through devices using KVM/QEMU, but I am not sure exactly if it would work down to such device level and specially on a MAC. It's normally...

Not possible with pure local storage, as it does not have any inbuilt replication between the two local storages. Few options. 1/ HA at the application level with 2 VM's one on each node. 2/ Use CEPH or another replicated storage system across the two nodes.

Then yes hammer is still supported by Proxmox.

They will be CEPH Hammer updates, you need to complete the following when you wish to update to Jewel : https://pve.proxmox.com/wiki/Ceph_Hammer_to_Jewel Which involves editing the sources to use jewel instead of hammer, when doing any for of update please make sure you use apt-get dist-upgrade

OVH

If your looking to add more than a minimum of : 1024 ceph osd pool set {pool-name} pg_num 1024 If you have no / minimal data then ceph osd pool set {pool-name} pgp_num 1024 If you have data in the pool then slowly increase pgp_num after each increase has taken effect.

You should never do an apt-get upgrade, always apt-get dist-upgrade If you try and run apt-get dist-upgrade this could/may fix your situation.

Correct, but if it's an empty pool (which this seems to be as only just got added correctly into Proxmox) it is easier/quicker to do.

Do you plan to add any more OSD's in the near future?