Well, the system has been designed to reduce network traffic wherever possible. That's the reason it also works reasonably well on low bandwidth networks like 64kbit (ISDN). I don't have any measurements but I think fast Ethernet (100Mbit) is more than enough. You won't see improvements when getting faster.
By sending SIGUSR and SiGUSR2 to the x2goagent you should get some internal performance counters. I don't remember what they show but it might help you getting an impression. Check nxagent's manpage for details.
Uli
Thanks for the info. And network hardware? The more gb transfer the
hardware supports, the better or is there a point where it doesn't
matter much anymore?
On 7/21/24 12:01 PM, Ulrich Sibiller wrote:
> Well,
>
> x2go compresses images using the CPU. So in theory the CPU affects the
> performance. And then you have the cache on disk. As that gets loaded
> on session startup/reconnect a very slow disk will increase the time
> for session startup and also for session suspension/shutdown when the
> cache is stored on disk (which might also happen during the session
> but I have never looked into that).
>
> However, these limitations are more or less theoretical. NX has been
> developed more than 20 years ago so on today's hardware the
> limitations are not relevant.
>
> Uli
>
> On Fri, Jul 19, 2024 at 12:47 AM sjomae <sjomae@mailbox.org> wrote:
>> Hi,
>>
>> I was wondering, what other hardware components besides network
>> hardware, determines the performance of x2go?
>>
>> Regards,
>>
>> /s
>>
>> _______________________________________________
>> x2go-user mailing list
>> x2go-user@lists.x2go.org
>> https://lists.x2go.org/listinfo/x2go-user