|
If you were logged in you would be able to see more operations.
|
|
|
|
Time Tracking:
|
|
Original Estimate:
|
1 hour
|
|
|
Remaining Estimate:
|
1 hour
|
|
|
Time Spent:
|
Not Specified
|
|
|
|
| Compatibility: |
backwards compatible
|
|
Thinking of how best to handle dynamic cluster deployment where no machine knows its hostname until deployed.
I've proposed a function to do this, a hostname() operation, but the other way is just to have the Hadoop Services explicitly do this -as an option.
something like :
conf.useLocalHostname
and we'd get the localhost (this could be cached somewhere BTW) and then fill in a new fs.default.name from the existing URL, replacing the hostname field.
|
|
Description
|
Thinking of how best to handle dynamic cluster deployment where no machine knows its hostname until deployed.
I've proposed a function to do this, a hostname() operation, but the other way is just to have the Hadoop Services explicitly do this -as an option.
something like :
conf.useLocalHostname
and we'd get the localhost (this could be cached somewhere BTW) and then fill in a new fs.default.name from the existing URL, replacing the hostname field. |
Show » |
|
Improvement
Resolved
Major
Datanodes and job tracker nodes to offer the ability to set their hostname dynamically, based on the machine doing the deployment
SFOS-1133is a better solution to this