Crucible1 #169
Crucible1 #169
Conversation
-should not be necessary if proper packages (lib-rdma) are installed -may not have ofed installed
-some people may miss it
There was a problem hiding this comment.
This looks largely benign. I would probably want to get rid of the removal of a bunch of debug lines that are commented out. When trying to trace what is happening those are strategically placed based on previous experience and they aren't really hurting anything.
| @@ -244,7 +244,7 @@ if [ -d ${trex_dir} -a -d ${tmp_dir} ]; then | |||
| else | |||
| vlan_opt="" | |||
| fi | |||
| trex_server_cmd="./t-rex-64 -i -c ${trex_cpus} --checksum-offload --cfg ${yaml_file} --iom 0 -v 4 --prefix trafficgen_trex_ ${vlan_opt}" | |||
| trex_server_cmd="./t-rex-64 -i -c ${trex_cpus} --no-ofed-check --checksum-offload --cfg ${yaml_file} --iom 0 -v 4 --prefix trafficgen_trex_ ${vlan_opt}" | |||
There was a problem hiding this comment.
So what is the side affect of this? I'm guessing Mellanox can't be used at all on the TRex side?
There was a problem hiding this comment.
No, this actually allows us to use Mellanox without TRex complaining.
| ip = ip + "." + str(octet) | ||
| return ip | ||
|
|
||
|
|
There was a problem hiding this comment.
You determined that this is backwards compatible, right?
| ip_src = { "start": ip_to_int(ip_src) + flow_offset, "end": ip_to_int(ip_src) + tmp_num_flows + flow_offset } | ||
| ip_dst = { "start": ip_to_int(ip_dst) + flow_offset, "end": ip_to_int(ip_dst) + tmp_num_flows + flow_offset } | ||
| ip_src = { "start": int_to_ip(ip_to_int(ip_src) + flow_offset), "end": int_to_ip(ip_to_int(ip_src) + tmp_num_flows + flow_offset) } | ||
| ip_dst = { "start": int_to_ip(ip_to_int(ip_dst) + flow_offset), "end": int_to_ip(ip_to_int(ip_dst) + tmp_num_flows + flow_offset) } |
There was a problem hiding this comment.
As noted above, I believe you found this is backwards compatible.
|
I'm a little confused by one aspect of this PR/branch. The first commit that I see (d18bbee) doesn't show up in the "Files Changed" diff -- or am I missing something? Make me wonder what else I may not be seeing... |
Never mind. I see that you removed this change with the last commit. |
changes to work with crucible