Thanks. It turns out that not all program use $http_proxy in bashrc. For
pip, which I had problem I had to use "pip --proxy http://somewhere install
pkg".
So, there is not problem with srun. Hope that it helps other too.
Regards,
Mahmood
On Mon, Aug 3, 2020 at 7:28 PM Matthew BETTINGER <
matt
Hello,
Not sure what your setup is but check compute nodes route table. Also might
need to turn on ipv4 forwarding on whatever is their default gw. Then also
firewalls can come in to play too. This isn't a slurm issue , pretty sure!
Matt
On 8/2/20, 7:53 AM, "slurm-users on behalf of Mahmo
Hi Mahmood,
I had the same problem some time ago when I took control of the cluster and
wanted to update the OS. While the login node has access to internet, the
calculation nodes are connected to the login node and is a private network.
Therefore, the only machine with access to the internet was
This is very likely by design of the cluster and/or network. Otherwise
users could use the cluster to mine bitcoin and such.
Brian Andrus
On 8/2/2020 7:11 AM, Mahmood Naderan wrote:
I thought that maybe srun doesn't transfer all settings from the head
node to the compute node.
The wget comman
I thought that maybe srun doesn't transfer all settings from the head node
to the compute node.
The wget command works on frontend but doesn't work on the compute.
mahmood@main-proxy:~$ wget google.com
--2020-08-02 16:05:55-- http://google.com/
Resolving google.com (google.com)... 216.58.215.238,
Probably unrelated to slurm entirely, and most likely has to do with
lower-level network diagnostics. I can guarantee that it’s possible to access
Internet resources from a compute node. Notes and things to check:
1. Both ping and http/https are IP protocols, but are very different (ping
isn’t
