Provided the fields really are in fixed positions, the unpack idea given by BrowserUK looks very apt. If that isn't the case (for example LONGERNAME can be instead of UPDATE) , I'd make a sequence of regexp rules to match on a what-comes-next basis and iterate on it, something like:

my $fieldSpec = [ { name => "name1", pattern => '\d{2}' },
                  { name => "trtype", pattern => '\D+' }, 
                  { name => "timestamp", pattern => '\d{8}' },
# etc.
];

my $results = [];

open my $fh, "<$file" or die $!;
while( <$fh> ) {
    my $result = {};
    for my $elt ( @$fieldspec ) {
        my $pat = $elt -> { pattern };
        /^($pat)(.*)$/ or die "unexpected format $_ at $.";
        $result -> { name } = $elt -> { name };
        $result -> { value } = $1;
        $_ = $2;
    }
    push @$results, $result;
}
[download]

...producing results as a reference to an array of hash in this particular example