gurubert's latest activity
-
gurubert replied to the thread Ceph will not mount after apt update.Updates should be made node for node and include a node reboot if necessary (new kernel, systemd etc). -
gurubert reacted to aaron's post in the thread 3rd Ceph MON on external QDevice (Podman) - 4-node / 2-site cluster with
Like.
For stretch PVE + Ceph clusters we recommend a full PVE install for the tie-breaker node. See the newly published guide: https://pve.proxmox.com/wiki/Stretch_Cluster -
gurubert reacted to alexskysilk's post in the thread How many nodes in a single cluster with Ceph requirement? with Like.
its an innocent enough question- but there are a lot of gotchas you need to consider. Clusters are made up of 3 elements- compute, storage, and networking. lets touch on each. COMPUTE: - Dell R630 is a 10 year old platform. as such, it offers...
-
Where did you get that idea?! ceph doesnt care about ANY of those things. Just be aware that the overall performance of the cluster will be as slow as the slowest monitor. ceph doesnt need or perform any live migration. As for memory, nodes...
-
gurubert reacted to MarkusKo's post in the thread Does Proxmox have a HCL (Hardware Compatibility List)? with Like.
Hi riahc3 There is no offical PVE HCL but PVE uses a modified Ubuntu Kernel you can check the Ubuntu HCL. PVE9 currently uses Kernel version 6.14 (Ubuntu HWE) and you can opt-in for a 6.17 Kernel. Just be aware that the Ubuntu HCL, as you...
-
gurubert reacted to anaxagoras's post in the thread Ceph keeps crashing, but only on a single node with Like.
I think you were correct. It hasn't crashed since i swapped out the entire system about 50 hours ago, i was getting daily crashes of ceph and hard lockups on the system...
-
gurubert reacted to alexskysilk's post in the thread Ceph - Reduced data availability: 3 pgs inactive with Like.
the ceph documentation has a dedicated page for pg troubleshooting: https://docs.ceph.com/en/latest/rados/troubleshooting/troubleshooting-pg/ this is also a good resource: https://ceph.io/geen-categorie/ceph-manually-repair-object/ generally...
-
gurubert replied to the thread Ethernet (ip a) how to figure out to which host they belong.The VMID or the CTID is the first number in the interface name. veth129i0 is the first interface of container 129.
-
gurubert reacted to Johannes S's post in the thread Neuling mit ersten - hoffentlich richtigen - Erkenntnissen --> ist das echt so? with Like.
Jein, die Standardsachen deckt die GUI ab, aber den kompletten Umfang was mit dem Linux-Unterbau möglich ( inklusive potentiell gefährlicher Aktionen) ist, kann und wird die GUI nie ganz abdecken. Eigentlich nicht...
-
gurubert replied to the thread Ceph - Reduced data availability: 3 pgs inactive.These PGs are lost if all of your OSDs are online again. You will have to remove them. Look in the Ceph documentation under PG Troubleshooting.
-
Tell Ceph to benchmark those drives again on OSD start and restart the service when appropriate: ceph config set osd osd_mclock_force_run_benchmark_on_init true There's also another ceph tell like command to run a benchmark right now, but I...
-
gurubert reacted to proxuser77's post in the thread [SOLVED] Does proxmox still offer a fully free version? with Like.
Because there's no need for such a plan when you can simply download the software for free and start using it. Would you really prefer having to go through a registration process, handing over your email address and personal data, only to get a...
-
gurubert reacted to proxuser77's post in the thread [SOLVED] Does proxmox still offer a fully free version? with Like.
Sorry, but you seem to be deliberately mixing things up, probably just because you’re frustrated that no one was able to help you in the other thread. With the Community Subscription, you’re paying for access to the Enterprise Repositories. The...
-
gurubert reacted to cosmos's post in the thread [SOLVED] Does proxmox still offer a fully free version? with Like.
@Ruprecht, I have been operating proxmox for 10 years without a subscription. Still using it today and I'm gonna buy the small business subscription for one of our work servers running proxmox. No, you don't have to buy it! You can stay on the...
-
Aber du kannst eine vm auch erstmal im .vmdk format auf einem pve benutzen (initial +last-sync) und später on-the-fly im pve zu raw/qcow2 umziehen. Bei einer Optimierung eines Vollumstiegs sollte man sich immer erstmal bewußt machen, wo der...
-
gurubert reacted to nl_alex's post in the thread [SOLVED] CEPH: public and cluster network with Like.
Hello Johannes, gurubert, thank you for your insights. I do have a third gigabit link reserved for corosync only, I probably should have mentioned that, but left it out since I only was interested in the ceph behaviour. Looks like I will run...
-
gurubert reacted to UdoB's post in the thread Proxmox VE with Ceph having PG warning after removing 3x OSD and rebuilt 3x OSD with Like.
That's just not how it works. "Failure domain" is host. You could have shutdown one host with all 4 OSDs going down. That would have worked fine. All data is distributed between those three nodes. You do not know onto which OSD. (Well, you...
-
gurubert reacted to robertlukan's post in the thread Proxmox VE with Ceph having PG warning after removing 3x OSD and rebuilt 3x OSD with Like.
Yes, most probably you have lost data. As you are having 3 copies of every PG, it can happen that PG was located on those three disks, so when you have removed them at the same time it became unavailable. As you have wiped OSDs, I dont think...
-
gurubert replied to the thread Proxmox VE with Ceph having PG warning after removing 3x OSD and rebuilt 3x OSD.You have lost data.
-
gurubert reacted to robertlukan's post in the thread Proxmox VE with Ceph having PG warning after removing 3x OSD and rebuilt 3x OSD with Like.
Have you waited for Ceph to recover after removing each OSD ?