Search results

rdma should be easy to implement in qemu initiator, it's just iser:// instead isci:// in the url. direct lun access through the kernel is not easy, because proxmox is distributed, and currently, if you remove a lun or resize it, or replace it (delete/add), the other nodes don't known about...

@bbgeek17 if you have time to test, I have send 2 patches: 1) bumping max qcow2 cache size up to 1GB memory (it should handle 8TB image without performance degradation) https://lists.proxmox.com/pipermail/pve-devel/2025-August/074555.html 2) saferemove speed improvement using blkdiscard...

could be interesting to see your custom crush rules

size=2/ min_size=2 . This is not going to work, it'll go readonly if you loose any osd. use min_size=1

Hi, I have already see this in production (on pve8), but I never found what was the problem. (So,I'm doing network config update carefully currently, with empty host, and reloading frr one by one). I wonder if it couldn't be related to the fwbr bridges for iptables , where frr could have...

if new kernel have changed interface name, the mtu from /etc/network/interfaces can't applied, as it's still defined on old name interface. (thanks systemd for the name change after each kernel update) pve9 have the new network interface pining to avoid this problem.

Fingerprints are only needed for self signed certicates, for normal certificate you really don't want to use fingerprint, or it'll break each the certificate is changed/refreshed.

@bbgeek17 in your article: "When a snapshot is created, subsequent writes are redirected to a new LVM logical volume. On a thin-provisioned SAN, storage for these writes is allocated on demand. When a snapshot is deleted, the system merges data from the previous QCOW image and LVM logical...

I have done test with blkdiscard --zeroing, from my tests, it's around 4~5x faster than the current csteam code (without throttling). Another benefit, is seem that is skip already zeroed block. so if you create,delete,create,delete,... it's a lot faster the second times if the block has not...

each snapshot have his own cache. (not that the cache value is a max value, by default 32MB max, but nothing is loaded until you begin to read differents part of disk). Also, metadatas not used are removed after 10min. I just sent a simple patch to the mailing, to increase the max cache size to...

about " When a snapshot is deleted, the system merges data from the previous QCOW image and LVM logical volume into the new QCOW image and logical volume. This merge process allocates additional storage in the new logical volume but does not release any storage from the old one." the option...

Hi, thanks for the article, I hope you enjoy my work ^_^ about performance, they are still performance improvement possible, mostly with qemu l2 cache size. Currently it's really depend of the qcow2 volume size, it's working up to around ~256GB. for bigger volume, we need to add an option in...

internet use 1500 mtu, so it's not going to work. At best, your router will refragment pakets to 1500. (you router lan need to be also 9000 btw)

not yet

That's seem pretty low. what is your cpu and what is your nic ? does the nic support vxlan offloading ?

when you edit /etc/pve/corosync.conf (in normal conditition, when you have quorum and write access to /etc/pve), the file is copied locally to each node in /etc/corosync/corosync.conf , then corosync is restarted. if you have done change ip or hostname change, and you don't have access to...

It's really looks like an hardware problem (maybe bad memory)/

The problem is, and I've read this somewhere else on the forums: that clustering only officially 100% works with nodes being on the same physical layer 2 network since corosync 3.X, it's use unicast udp (before it was multicast), so it's working on layer3 network. (I'm using it in a full l3...

can you try to add issue_discards=1 in lvm.conf ? (I'm sure it's working at lvremove , I don't known if it's mandatory when discard send by the guest os) Do you have good performance with ocfs2 ? Each time I'm trying I have pretty bad performance without preallocation full. and internal...

another thing not yet implemented is thin provisioning, where it'll be possible to have a smaller lvm volume than the defined size, with auto-extend of lvm volume. So, currently, each snapshot use the same size than the base lvm volume