czechsys's latest activity

  • C
    czechsys replied to the thread Help with cluster network optimization.
    PVE/ssh by default listen on all interfaces, so it's possible to connect via mgmt, corosync, storage etc IPs. If you don't need very high performance (100 Gbps networks), i will use 9000 only on limited scope, for example, dedicated vmbr/vlan for...
  • C
    czechsys replied to the thread 3 servers, 3 cables, 1 ceph network?.
    Depends on nic utilization, but we use the switch way for our 3 nodes cluster: 1x lacp (2 ports) with vlans for management (= ceph public), for vm, for corosync, etc 2x lacp (2 ports) with vlans for corosync, for ceph storage Mesh is for small...
  • C
    czechsys replied to the thread Second vmbr and NIC defeats the first in terms of WebUI, SSH.
    I am not using such config variant, so i can theoretize it looks ok. But i am using vlans everywhere and never assign ip to bridge, but using subinterface every time. Anyway, PVE can access multiple networks without fw/router. For nfs access you...
  • C
    czechsys replied to the thread bond health check.
    options: script it use mlag switches multipath etc etc
  • C
    czechsys replied to the thread Outgoing mail from internal servers rejected with “Relay access denied” in PMG.
    PMG is mainly for mailserver to mailserver communication. If you are trying sending mail from non mailservers, send those mails to exchange first. Or https://www.postfix.org/SMTPD_ACCESS_README.html#relay and test.
  • C
    czechsys replied to the thread [SOLVED] Hardening SSH.
    PVE require root via ssh for cluster function.
  • C
    czechsys replied to the thread Hardware advice (or "questionable Proxmox performance on nice box").
    Disabled power saving states on the Dell? Firmware updated? SSDs are enterprise or desktop versions? I have feeling, you have disk problem, missing virtio drivers in Window VMs etc, but nothing concrete, no VM config/pve versions written. Maybe...
  • C
    czechsys replied to the thread Test migration stuck.
    MTU is 1500 for all servers, switches has 9216. Remotes the same location, the same switch and vlan. Because internal CA, fingeprints (FP): Configured for REMOTE1:8006: from pveproxy-ssl.pem FP - BUT - it automatically detects intermediate...
  • C
    czechsys replied to the thread Test migration stuck.
    Nothing spotted yet, because remotes are in production and there is nothing non-standard even on netdata (i can miss something of course). Today 4x test moving real vm 10+200 GB, migration fails for 2nd disk around 17-19/16-17 GB position...
  • C
    czechsys replied to the thread Ceph Cluster Broken.
    Fix your time synchronization as the first step. clock skew detected on mon.proxmox03, mon.proxmox01
  • C
    czechsys replied to the thread Test migration stuck.
    Tests 10 GB new VM (not recreated between this test round, the same VMID): 1. slowed down 598585344 bytes (8.6 GB, 8.0 GiB) copied, 39 s, 220 MB/s 8803647488 bytes (8.8 GB, 8.2 GiB) copied, 40 s, 220 MB/s 8878555136 bytes (8.9 GB, 8.3 GiB)...
  • C
    czechsys replied to the thread iSCSI SAN Configuration.
    https://pve.proxmox.com/wiki/Multipath
  • C
    czechsys replied to the thread Test migration stuck.
    Created test VM1 with 10 GB and VM2 with 10+10 GB without any OS. Tested PDM migration: 1st try: VM1 slowed to hell 2nd try: VM2 succeeded Remotes tests: SERVER1: if=/dev/pve/vm-104-disk-1 bs=64k status=progress of=/dev/null - skipped due...
  • C
    czechsys replied to the thread Test migration stuck.
    Both sides are LVM thick (HP G8 servers with Pxxx smartaray raid10) src: 4 0 1573 1 20 0 179628 137020 hrtime Ss ? 272:12 pvestatd 0 0 1593 1 20 0 219244 175984 hrtime Ss ? 0:31 pvedaemon 1...
    • Screenshot from 2025-09-19 13-54-39.png
  • C
    czechsys replied to the thread Test migration stuck.
    The same test as with Alpha version. The same remotes, upgraded to PVE 9. Migration of the 2nd disk stucked, manually stopped. Remotes: pve-manager/9.0.6/49c767b70aeb6648 (running kernel: 6.14.11-1-pve) 2025-09-19 11:28:53 remote: started...
  • C
    czechsys replied to the thread Mark found message as spam via webGUI?.
    That's not possible, because message doesn't exists on PMG after delivering to mailserver. So nothing to scan/learn again.
  • C
    czechsys replied to the thread Is clustering production-ready on Proxmox VE 8.2.2?.
    Don't use 8.2.2. You will have problem install another host with this version. Use last PVE version, now 8.4, or go directly to 9.
  • C
    czechsys replied to the thread A few questions about Backup.
    Filebackup of the host. Especially /etc/pve for pve specific configuration, large part of it is handled by cluster, but something is node dependent (example: certificates). Search forum at least.
  • C
    czechsys replied to the thread New to ProxMox and stabillity.
    Running PVE 9 on DL360/380 Gen 8-10 without problem. You can install rsyslog, attach monitor for kernel dump, maybe there will be something not in journal. It's standard debian anyway. Other thing - firmware, power, etc etc etc.
  • C
    czechsys replied to the thread Best logs to collect on Proxmox when LVM volumes disappear due to SAN/FC issues.
    Without upgrading the 2nd SAN node: What will happen, when you shutdown the 2nd SAN node? What will happen, when you shutdown the 1st SAN node? I think, it's SAN support thing mainly, because with 2nd SAN node upgrading it blocks something...
Show older items
Top Bottom