Naming of Key Pairs
Closed this issue · 1 comments
Hi,
I have used the Create Key pair(s)
option from the Deploy
menu. This resulted in the creation of a private key Matrixuseast1-1
in ~/Keys
where the filename has suffix -1
.
MATRIX/Deployment/aws_deploy.py
Lines 29 to 30 in 77a3316
During deployment, the Matrix client tries to use the key name from regions.json
where this suffix is omitted (it is just Matrixuseast1
).
MATRIX/Deployment/aws_deploy.py
Lines 116 to 132 in 77a3316
Choose deployment task
1. Deploy Instance(s)
2. Create Key pair(s)
3. Create security group(s)
4. Get instances network data
5. Terminate machines
6. Change machines types
7. Start instances
8. Stop instances
Your choice:1
Current date :
2018-09-11 14:23:54.714932
Deploying instances :
region : us-east-1b
number of instances : 3
ami_id : ami-08c9a677
instance_type : c4.large
valid until : 2018-09-11 20:23:54.714932
THE KEY PAIR 'MATRIXUSEAST1' DOES NOT EXIST
Waiting for the images to be deployed..
I can change the entry of the regions.json
, but I am not sure what happens when there are multiple regions involved.
Hi,
At your config file, you defined your requested regions:
MATRIX/ProtocolsConfigurations/Config_ABY_InnerProduct.json
Lines 30 to 33 in 77a3316
The
b
at the end specify the placement group inside the region.Since the
regions.json
file is common to all the placement groups inside a region, the code snippet omits the placemant group specifier and not the suffix of the key name.
Hence you should change the regions.json
file according to your key file name.