Differences

This shows you the differences between two versions of the page.

Link to this comparison view

irc:1446591600 [2017/05/27 13:44] (current)
Line 1: Line 1:
 +[01:05:19] <​DP_2014>​ I think there is a bug in io.vertx.ext.web.impl.Locale
 +
 +[01:06:50] <​DP_2014> ​ static default Locale create(String language, String country) {        return new LocaleImpl(language,​ country, (String)null); ​   }     ###### ​ (String)null ​    sets the variant to null value, ​ which when   it gets converted to  java.util.Locale ​ throws an  Exception ​ as variant cannot be null.
 +
 +[09:15:19] *** ChanServ sets mode: +o purplefox
 +
 +[12:45:17] *** ChanServ sets mode: +o temporalfox
 +
 +[14:44:20] *** ChanServ sets mode: +o temporalfox
 +
 +[14:57:45] <​tsegismont>​ Hi everyone
 +
 +[14:58:15] <​tsegismont>​ I'm working on NetClient and HttpClient metrics support in vertx-monitor
 +
 +[14:58:34] <​tsegismont>​ There'​s one thing which is a bit annoying
 +
 +[14:59:34] <​tsegismont>​ When you create a client connection with, for example, host=mydomain.com
 +
 +[14:59:51] <​tsegismont>​ in the io.vertx.ext.hawkular.impl.NetClientMetricsImpl#​connected callback
 +
 +[15:00:24] <​tsegismont>​ you don't get a remoteAddress parameter with host=mydomain.com,​ but the actual IP
 +
 +[15:01:22] <​tsegismont>​ So it seems difficult, if not impossible, to implement host-based filtering
 +
 +[15:01:54] <​tsegismont>​ For example: I want to configure vertx-monitor to send metrics for TCP connections to mydomain.com only
 +
 +[15:02:01] <​tsegismont>​ Should I open an issue?
 +
 +[16:15:49] *** ChanServ sets mode: +o temporalfox
 +
 +[16:19:12] <​tsegismont>​ temporalfox,​ hi, have you seen the comments above?
 +
 +[16:19:21] <​temporalfox>​ nope, I was away
 +
 +[16:19:25] <​temporalfox>​ ah
 +
 +[16:19:27] <​temporalfox>​ actually yes
 +
 +[16:19:31] <​temporalfox>​ at 3pm ?
 +
 +[16:19:34] <​tsegismont>​ yep
 +
 +[16:19:52] <​temporalfox>​ there is a DNS client you can use
 +
 +[16:21:44] <​tsegismont>​ temporalfox,​ that wouldn'​t work
 +
 +[16:21:55] <​tsegismont>​ if you connect to www.google.fr
 +
 +[16:22:05] <​tsegismont>​ it will be resolved to an IP
 +
 +[16:22:13] <​tsegismont>​ but then if you lookup this IP
 +
 +[16:22:27] <​tsegismont>​ you will get something like lhr08s07-in-f35.1e100.net
 +
 +[16:22:36] <​tsegismont>​ doesn'​t match
 +
 +[16:23:19] <​temporalfox>​ so you want this info in the net socket metrics client callback ?
 +
 +[16:23:45] <​tsegismont>​ yes, so that in metrics config object
 +
 +[16:23:53] <​tsegismont>​ the user can say
 +
 +[16:24:18] <​temporalfox>​ so it's a metrics issue and not really a netsocket client issue right ?
 +
 +[16:24:34] <​tsegismont>​ "I want to monitor http client connections to api.anawesomeservice.com"​
 +
 +[16:24:49] <​tsegismont>​ temporalfox,​ yes and no
 +
 +[16:24:50] <​tsegismont>​ :)
 +
 +[16:25:10] <​temporalfox>​ perhaps that the client could resolve the name of api.anawesomeservice.com
 +
 +[16:25:12] <​tsegismont>​ The Metrics API gives you a remote address (SocketAddress)
 +
 +[16:25:17] <​temporalfox>​ and filter according to the IP
 +
 +[16:25:25] <​temporalfox>​ I mean the metrics spi impl
 +
 +[16:25:33] <​temporalfox>​ just a thought
 +
 +[16:25:33] <​tsegismont>​ but it does so querying the netty channel
 +
 +[16:25:45] <​tsegismont>​ instead of the parameters which were used to create the connection
 +
 +[16:25:54] <​temporalfox>​ please open an issue anyway and we'll figure out
 +
 +[16:26:07] <​temporalfox>​ in the vert-x3/​issues or in bugzilla eclipse
 +
 +[16:26:26] <​tsegismont>​ see io.vertx.core.net.impl.NetClientImpl#​connected
 +
 +[16:26:43] <​tsegismont>​ sock.setMetric(metrics.connected(sock.remoteAddress()));​
 +
 +[16:27:11] <​tsegismont>​ io.vertx.core.net.impl.ConnectionBase#​remoteAddress
 +
 +[16:30:13] <​tsegismont>​ temporalfox,​ will do
 +
 +[17:01:59] <​stampy88>​ hi all. I was hoping that someone would have a chance to review this PR https://​github.com/​eclipse/​vert.x/​pull/​1190
 +
 +[17:02:19] <​stampy88>​ I originally submitted it a month ago, but re-submitted last night because the IP Validation was not passing
 +
 +[20:24:43] <​Kioz>​ Hi
 +
 +[20:56:47] <​stampy88>​ pmlopes: if you have time, can you take a look at this PR? https://​github.com/​eclipse/​vert.x/​pull/​1190
 +
 +[20:58:22] <​stampy88>​ this is a painful to code around now with try..catches