Latest activity

What's new Featured content New posts New profile posts Latest activity
  • Y
    yummiweb replied to the thread [BUG?] PBS Sync (pull) completely ignores bandwidth setting.
    Thanks for the heads-up! I've noticed that myself in the meantime. It might be interesting to understand why I didn't see any speed difference in my tests: This was probably due to sshuttle, which in this configuration (standard buffer or no...
  • E
    Erik Horn posted the thread rp_filter changes in PVE 9 in Proxmox VE: Networking and Firewall.
    I've recently started testing the upgrade of our v8.4 cluster to v9.1 (patched using the nosub repository today). While troubleshooting issues with our EVPN configuration, I found that the previous method for disabling reverse path filtering...
  • Y
    yummiweb replied to the thread [BUG?] PBS Sync (pull) ignoriert Bandbreiteneinstellung komplett.
    Und noch ein Nachtrag - und wohl die Lösung Ich habe die "--no-latency-control" Option wieder aus dem sshuttle Befehl entfernt und die sshuttle verbindung neu gestartet, seitdem läuft der job mit einer Geschwindigkeit von ca. 40 Mbits (bzw...
  • M
    matada replied to the thread SATA link down issues.
    So it's not just me with this issue. Well that's good to know I guess? I thought the single hdd in my server was dying because it'd randomly disappear and crash any containers that used the NFS mount for it when there was heavy read/write to...
  • Y
    yummiweb replied to the thread [BUG?] PBS Sync (pull) ignoriert Bandbreiteneinstellung komplett.
    Nachtrag: Per rsync von derselben PBS Maschine (mit dem ich auch Dateibackups offsite sichere) kann ich die Bandbreite sauber nach oben und unten bestimmen - es liegt also nicht an irgendwelchen anderen Netzwerkbeschränkungen. Rsync uns Sshuttle...
  • J
    jauling reacted to iscrow's post in the thread CPU with AVX support, with Like Like.
    My solution was to define a new CPU model based on x86-64-v2-AES that also exposes AVX New models can be defined here cluster wide: /etc/pve/virtual-guest/cpu-models.conf Add this definition to the /etc/pve/virtual-guest/cpu-models.conf file...
  • Y
    yummiweb replied to the thread [BUG?] PBS Sync (pull) ignoriert Bandbreiteneinstellung komplett.
    sync: s-1a22b33c-4dd0 comment Zeitgleiche sshuttle Verbindung erforderlich, siehe crontab Eintrag encrypted-only false ns owner pbs_user_org@pbs rate-in 16 MiB remote PBS_ORG remote-ns...
  • B
    buggy09 replied to the thread 2 Node with Msa.
    Few weeks ago we just completed migration of 20-ish VMs (mostly windows, few linux) from ESX/Veeam to Proxmox PVE/PBS using similar environment as yours (2 HPE Proliant hosts and MSA 2060 FC as shared storage). Everything works perfect. You...
  • R
    Raudi replied to the thread Absturz oder Reboot verbleibender Hosts bei Update und reboot des Quorum.
    Was mich wundert warum sollte HA hier die Hosts rebooten? Wenn 2 verbleibende Hosts aktiv sind und einer fehlt, dann dann müssen doch die 2 verbleibenden merken, dass sie aktiv sind und einer gestorben ist. Mar 06 20:04:28 pve01 corosync[1858]...
  • S
    snowa reacted to Wofferl's post in the thread [SOLVED] 400 Bad Request on VM Backup with Like Like.
    To explain what the problem here is, now that it has been found (400 Bad Request on raspberry pi). The raspi 5 has by default a kernel with 16k pagesize, this probably leads to a problem when writing or creating the fixed index, which leads to...
  • SteveITS
    SteveITS replied to the thread Ceph rbd du shows usage 2-4x higher than inside VM.
    One of the smaller VMs is off so I ran rbd sparsify and it worked. The syntax I found was a bit different, rbd sparsify --pool poolname diskname, but it did immediately reduce the rbd du usage from 11 GB to 8.8 GB, and only took a few seconds...
  • C
    cliffpercent posted the thread Long-time offline of a node in Proxmox VE: Installation and configuration.
    I have a node with hardware failure, its resurrection has been deferred repeatedly now (soon a year). HA votes are set to 0, quorum and capacity is OK with the remaining odd number of nodes. The rest of the nodes have been receiving updates as...
  • A
    anotherhaus reacted to ribbon's post in the thread [SOLVED] token already exists with Like Like.
    Solved! 1) Visit the target PVE. 2) Select Datacenter. 3) Select Permission->API Token. 4) You will then find the API Token that PDM created. Delete it.
  • B
    Bu66as reacted to TErxleben's post in the thread OpenWRT-Gateway Failover per VM with Like Like.
    Habe ich mal umgebaut und werde es beobachten. Geändertes checkwan.sh: #!/bin/bash # echo 0 #curl -s https://icanhazip.com | wc -l # HOPS=$(traceroute -4 8.8.8.8 | wc -l) # if [ $HOPS -gt 5 ]; then # echo 1 # else # echo 0 # fi ping -c1 -W3...
  • J
    jaminmc replied to the thread [TUTORIAL] Custom Proxmox 6.19.0 Kernel. 6.18.0 is also available for Testing..
    I have 2 new releases. One is proxmox-kernel 6.19.2 (unofficial) — Linux 6.19.2 + OpenZFS 2.4.1, which is based on Ubuntu 6.19.0-6.6. I made it an official release, as The Resolute Raccoon has it as a stable release. I have updated it to have...
  • J
    Johannes S reacted to bbgeek17's post in the thread LVM-Thin oder ZFS with Like Like.
    Bieten Sie die Vorteile von Berliner vs. Pfannkuchen vs. Krapfen Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
  • N
    networkguy3 replied to the thread [SOLVED] Erasure Code Pool WAL and RocksDB usage.
    This is shared storage, so would you suggest mounting the storage in a VM and sharing it with Samba or sharing it with SMB from ZFS itself?
  • B
    beisser reacted to Bu66as's post in the thread OpenWRT-Gateway Failover per VM with Like Like.
    Stimmt, wenn beide VMs die gleiche MAC haben fällt das ARP-Thema komplett weg, guter Move. Dann vergiss den arping-Tipp. Einzige Sache: falls durch nen Bug oder Race Condition doch mal beide VMs kurz gleichzeitig laufen (start/shutdown ist ja...
  • gurubert
    gurubert reacted to BD-Nets's post in the thread OCFS2(unsupported): Frage zu Belegung with Like Like.
    Prüft diesen Plan bitte gründlich. Wie schon geschrieben bin ich eher auf den Nachfolgern (3par) zu Hause, aber da gibt es folgende Fallen: Die Platten haben eine spezielle Formatierung (520 statt 512 Byte/Sektor) Die Platten brauchen spezielle...
  • gurubert
    gurubert reacted to Bu66as's post in the thread OCFS2(unsupported): Frage zu Belegung with Like Like.
    Die "1 MiB Cluster = Ursache" Story stimmt so nicht. Cluster-Overhead ist real, aber das sind ein paar Prozent, keine 2+ TiB. Das Problem bleibt der Controller der die UNMAPs nicht umsetzt, das ist unabhängig vom Filesystem. Was @Johannes S zum...
Show older items
Top Bottom
Back