[pve-devel] [PATCH docs] storage: note that qcow2 internal snapshots are inefficient

[pve-devel] [PATCH docs] storage: note that qcow2 internal snapshots are inefficient

[Fiona Ebner]

It's a commonly reported issue, most recently again in the enterprise
support, that taking or removing snapshots of large qcow2 files on
file-based network storages can take a very long time. Add a note
about this limitation.

Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
 pvesm.adoc | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/pvesm.adoc b/pvesm.adoc
index d36baf8..165b446 100644
--- a/pvesm.adoc
+++ b/pvesm.adoc
@@ -88,6 +88,11 @@ block device functionality.
 
 ^2^: On file based storages, snapshots are possible with the 'qcow2' format,
 either using the internal snapshot function, or snapshots as volume chains^4^.
+Creating and deleting internal 'qcow2' snapshots will block a running VM and
+is not an efficient operation. The performance is particularly bad with network
+storages like NFS. On some setups and for large disks (multiple hundred GiB or
+TiB sized), these operations may take several minutes, or in extreme cases, even
+hours.
 
 ^3^: It is possible to use LVM on top of an iSCSI or FC-based storage.
 That way you get a `shared` LVM storage
-- 
2.47.3



_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel

Re: [pve-devel] [PATCH docs] storage: note that qcow2 internal snapshots are inefficient

[Fiona Ebner]

Am 28.11.25 um 4:56 PM schrieb Fiona Ebner:
> It's a commonly reported issue, most recently again in the enterprise
> support, that taking or removing snapshots of large qcow2 files on
> file-based network storages can take a very long time. Add a note
> about this limitation.
> 
> Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
> ---
>  pvesm.adoc | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/pvesm.adoc b/pvesm.adoc
> index d36baf8..165b446 100644
> --- a/pvesm.adoc
> +++ b/pvesm.adoc
> @@ -88,6 +88,11 @@ block device functionality.
>  
>  ^2^: On file based storages, snapshots are possible with the 'qcow2' format,
>  either using the internal snapshot function, or snapshots as volume chains^4^.
> +Creating and deleting internal 'qcow2' snapshots will block a running VM and
> +is not an efficient operation. The performance is particularly bad with network
> +storages like NFS. On some setups and for large disks (multiple hundred GiB or
> +TiB sized), these operations may take several minutes, or in extreme cases, even
> +hours.
>  
>  ^3^: It is possible to use LVM on top of an iSCSI or FC-based storage.
>  That way you get a `shared` LVM storage

Ping

Re: [pve-devel] [PATCH docs] storage: note that qcow2 internal snapshots are inefficient

[Maximiliano Sandoval]

Fiona Ebner <f.ebner@proxmox.com> writes:

> Am 28.11.25 um 4:56 PM schrieb Fiona Ebner:
>> It's a commonly reported issue, most recently again in the enterprise
>> support, that taking or removing snapshots of large qcow2 files on
>> file-based network storages can take a very long time. Add a note
>> about this limitation.
>> 
>> Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
>> ---
>>  pvesm.adoc | 5 +++++
>>  1 file changed, 5 insertions(+)
>> 
>> diff --git a/pvesm.adoc b/pvesm.adoc
>> index d36baf8..165b446 100644
>> --- a/pvesm.adoc
>> +++ b/pvesm.adoc
>> @@ -88,6 +88,11 @@ block device functionality.
>>  
>>  ^2^: On file based storages, snapshots are possible with the 'qcow2' format,
>>  either using the internal snapshot function, or snapshots as volume chains^4^.
>> +Creating and deleting internal 'qcow2' snapshots will block a running VM and
>> +is not an efficient operation. The performance is particularly bad with network
>> +storages like NFS. On some setups and for large disks (multiple hundred GiB or
>> +TiB sized), these operations may take several minutes, or in extreme cases, even
>> +hours.
>>  
>>  ^3^: It is possible to use LVM on top of an iSCSI or FC-based storage.
>>  That way you get a `shared` LVM storage
>
> Ping

In my experience, users mostly into issues when deleting qcow2 snapshots
on NFS can take up to 10 hours. I would personally put more emphasis on
this specific combination being problematic, and probably mention that
in such case one should delete snapshots while the VM is offline.

-- 
Maximiliano

Re: [pve-devel] [PATCH docs] storage: note that qcow2 internal snapshots are inefficient

[Fiona Ebner]

Am 23.03.26 um 10:58 AM schrieb Maximiliano Sandoval:
> Fiona Ebner <f.ebner@proxmox.com> writes:
>> Am 28.11.25 um 4:56 PM schrieb Fiona Ebner:
>>> It's a commonly reported issue, most recently again in the enterprise
>>> support, that taking or removing snapshots of large qcow2 files on
>>> file-based network storages can take a very long time. Add a note
>>> about this limitation.
>>>
>>> Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
>>> ---
>>>  pvesm.adoc | 5 +++++
>>>  1 file changed, 5 insertions(+)
>>>
>>> diff --git a/pvesm.adoc b/pvesm.adoc
>>> index d36baf8..165b446 100644
>>> --- a/pvesm.adoc
>>> +++ b/pvesm.adoc
>>> @@ -88,6 +88,11 @@ block device functionality.
>>>  
>>>  ^2^: On file based storages, snapshots are possible with the 'qcow2' format,
>>>  either using the internal snapshot function, or snapshots as volume chains^4^.
>>> +Creating and deleting internal 'qcow2' snapshots will block a running VM and
>>> +is not an efficient operation. The performance is particularly bad with network
>>> +storages like NFS. On some setups and for large disks (multiple hundred GiB or
>>> +TiB sized), these operations may take several minutes, or in extreme cases, even
>>> +hours.
>>>  
>>>  ^3^: It is possible to use LVM on top of an iSCSI or FC-based storage.
>>>  That way you get a `shared` LVM storage
>>
>> Ping
> 
> In my experience, users mostly into issues when deleting qcow2 snapshots
> on NFS can take up to 10 hours. I would personally put more emphasis on
> this specific combination being problematic, and probably mention that
> in such case one should delete snapshots while the VM is offline. 

NFS is already explicitly mentioned as being particularly bad. We also
had reports where creating snapshots on NFS took hours.

Doing it offline doesn't change the fact that it takes that long, but I
guess doing it while offline is still better, because being paused for a
long time is not that nice from a guest perspective.

I'll add the following in v2:
"If your setup is affected, create and remove snapshots while the VM is
shut down, expecting a long task duration."