Increase retry attempts when downloading service yaml files

[niusmallnan]

RancherOS Version: (ros os version)
v1.5.4

Where are you running RancherOS? (docker-machine, AWS, GCE, baremetal, etc.)
ALL

From netconfig.log it is shown that the nameserver record was written at 12:47:47:

time="2019-12-05T12:47:47Z" level=info msg="writing to /etc/resolv.conf: nameservers: [192.168.200.1 192.168.200.2], search: []" 
time="2019-12-05T12:47:47Z" level=debug msg="Resolve.conf == [nameserver 192.168.200.1
nameserver 192.168.200.2
], <nil>" 

This means that the resolution of domain example.com could not work before ​12:47:47, but the logs also show RancherOS trying to download service yaml files at this time.

time="2019-12-05T12:47:45Z" level=debug msg="start trying LoadFromNetwork(http://example.com/rancher/os-services/v1.5.4/o/open-vm-tools.yml)" 
time="2019-12-05T12:47:45Z" level=debug msg="Fetching data from http://example.com/rancher/os-services/v1.5.4/o/open-vm-tools.yml. Attempt #1"
...
...
time="2019-12-05T12:47:47Z" level=error msg="failed to LoadFromNetwork: Unable to fetch data. Maximum retries reached: 3" 
time="2019-12-05T12:47:47Z" level=error msg="Unable to fetch data. Maximum retries reached: 3" 

We should increase the default number of retries, which should mitigate this problem.

[kingsd041]

In rancheros v1.5.5-rc1, we have added the rancher.http_load_retries: 6 parameter to resolve this issue