Github user SolidWallOfCode commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/240#discussion_r33824514
--- Diff: iocore/hostdb/HostDB.cc ---
@@ -2771,11 +2586,14 @@ ParseHostLine(char *l)
int n_elts = elts.Initialize(l, SHARE_TOKS);
// Elements should be the address then a list of host names.
// Don't use RecHttpLoadIp because the address *must* be literal.
- HostFilePair item;
- if (n_elts > 1 && 0 == item.ip.load(elts[0]) && !item.ip.isLoopback()) {
+ IpAddr ip;
+ if (n_elts > 1 && 0 == ip.load(elts[0]) && !ip.isLoopback()) {
for (int i = 1; i < n_elts; ++i) {
- item.name = elts[i];
- HostFilePairs.push_back(item);
+ ts::ConstBuffer name(elts[i], strlen(elts[i]));
+ // If we don't have an entry already (host files only support single
IPs for a given name)
--- End diff --
Is this true? I've done quite a bit of research on this and can't verify
that accuracy of that claim. I wrote the previous version on the presumption
you could have round robin data for a host by using multiple lines with
distinct IP addresses and the same FQDN.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
