You can read the first line using the regular method - open, read to first record delimiter. Then just split to get your field names, and convert them to a hash so you can find the proper field in later records. That's no problem, even if fields are added or removed.
chomp($record = <DATA>);
$keys{$_} = $c++ for split / /, $record;
## Get last record and put it in $record
@record = split / /, $record;
print $record[$keys{'timestamp'}];
The harder part is getting at the last record. For this, jump to the end of the file minus whatever the maximum possible record size could be (+ record delimiter), then read everything from there to the end and retrieve the last record. Again, not that difficult.