gurubert's latest activity
-
systemd-userdbd where the change( adding an optional field you can set to any date you like) occured isn't even used in ProxmoxVE. But even if it would be used: The whole outcry is way overblown. Even if one get rids of systemd ( imho a bad idea... -
Hey all. We have put together a storage plugin for S3 storage for PVE. This comes from us having used a local directory and s3fs (fuse) with mixed results - bad things happen when the S3 storage goes offline, either via a proxy outage or...
-
gurubert replied to the thread OCFS2(unsupported): Frage zu Belegung.Warum sollte Multipathing den discard-Befehl nicht weitergeben? Bzw sollte das Discard ja die qcow2-Datei im OCFS2 kleiner machen. Auf der Ebene hat das mit Multipathing zum SAN ja noch gar nichts zu tun. Wenn das SAN-Storage selber die LUN...
-
@spirit: Thank you for the concrete numbers and references — that's very helpful. I was wrong to be dismissive earlier about the allocator angle. The Ceph blog you linked (QEMU/KVM Tuning) reports a ~50% improvement on 16KB random reads (53.5k →...
-
gurubert reacted to Astrodata's post in the thread Proxmox's scheduling syntax does something weird with
Like.
thank you for your clarification. little bit unpractical that the event is beeing skipped and not beeing executed after the time change. but mostly no big deal to avoid the problem with setting another time. even that this is not a pbs related... -
gurubert reacted to Onslow's post in the thread Proxmox's scheduling syntax does something weird with Like.
This is another example of systemd "Knows Better" dogma :-(. Cron is more sensible: "Daylight Saving Time and other time changes Local time changes of less than three hours, such as those caused by the start or end of Daylight Saving Time, are...
-
gurubert reacted to Onslow's post in the thread Proxmox's scheduling syntax does something weird with Like.
Welcome, @Astrodata I have the same output in a system which has nothing common to Proxmox: @kubuntu:~$ systemd-analyze calendar "Sun *-*-1..31 02:00:00" Original form: Sun *-*-1..31 02:00:00 Normalized form: Sun *-*-01..31 02:00:00 Next...
-
gurubert reacted to VictorSTS's post in the thread Proxmox/Ceph - Disable OSD rebalancing with Like.
In a 3 node cluster with size=3, min_size=2 and the default replica rule, Ceph won't rebalance anything if one node fails. To comply with the default rule "three replicas en three OSD located in three different servers" you need 3 servers, if...
-
gurubert reacted to lucavornheder's post in the thread Configure migration settings network to share the same network of Ceph with Like.
Yeah, you can switch the network in the Datacenter menu under Options. Just keep in mind that 2x10G isn’t a lot for Ceph, if a migration starts running, you can hit bottlenecks pretty fast. I think, there’s also an option to limit the migration...
-
gurubert reacted to LnxBil's post in the thread Proxmox seems to use outdated cloud-init Module with Like.
The package cloud-init is part of Debian. This version normally fixed for each Debian release, so with the advent of Debian 14 Forky and upcoming PVE 10 the current version available right now is 25.3, so not that much newer. That may for Debian...
-
gurubert reacted to Alwin Antreich's post in the thread Ceph erasure coding configuration best practices with Like.
What is FTT in this context? The memory target for OSDs should be 8 GB, which means 8GB per daemon. Anything lower usually sacrifices performance during recovery. With the memory target using EC or replica doesn't make a difference. Yes...
-
gurubert replied to the thread Ceph - VM with high IO wait.By default /etc/ceph/ceph.conf is a symlink to /etc/pve/ceph.conf which makes it the same on each cluster node. AFAIK it is easier to use ceph config set to set the values in the config db for each Proxmox node. ceph config set client.HOSTNAME...
-
gurubert replied to the thread OCFS2(unsupported): Frage zu Belegung.Würde ich so auch vermuten.
-
gurubert reacted to alexskysilk's post in the thread Make a Proxmox NTFS HDD accessible under Windows network (Windows Explorer) with Like.
I misspoke. PVE does not have any such functionality. What I MEANT was to do it at the local debian level (and since thats where the entirety of the pve stack resides it was convenient shorthand) there are a bunch of tutorials for "setting up...
-
gurubert reacted to dietmar's post in the thread Age verification is identity verification - archLinux32 banned in Brazil. with Like.
Proxmox VE is a bare-metal hypervisor platform aimed at IT professionals and system administrators. It is not a product or service directed at children ...
-
gurubert reacted to Johannes S's post in the thread Einführung von Einschränkungen / Kennzeichnungspflicht von KI-Beiträgen with Like.
Überrannt trifft es ziemlich gut, wenn Leute anstatt mal versuchen ihr Problem zu verstehen und mit eigenen Worten zu beschreiben sich das durch die KI vorformulieren lassen und nicht mal mehr checken, ob das überhaupt ihr Problem richtig...
-
gurubert reacted to proxuser77's post in the thread Einführung von Einschränkungen / Kennzeichnungspflicht von KI-Beiträgen with Like.
Da bin ich deiner Meinung. Ich nutze sie beispielsweise für Rechtschreibprüfungen bei längeren Texten und für Übersetzungen bei englischen Beiträgen. Manchmal nutze ich sie auch zur Ideensammlung oder um erste Abklärungen zu treffen, was das...
-
Looking at the graphs you posted, a few things stand out. The Ceph pool metrics show almost no writes — IOPS and throughput on `pve_ceph_prod_3az` are nearly 100% read. Cross-AZ write latency is not what's hurting you here. The good news is that...
-
gurubert reacted to Marco83's post in the thread Proxmox SDN Traffic breakout Interface and routing with Like.
I was thinking abaou PBR (policy-based routing)—i.e., forcing SDN traffic out through a different NIC. Right now, in my cluster the management VLAN/interface has the default route, which means the SDN traffic would break out into my management...
-
gurubert replied to the thread Ceph - VM with high IO wait.What is the network latency between the nodes, especially between the different AZs?