Summary: | Frequent timeouts at jabber.org | ||
---|---|---|---|
Product: | Telepathy | Reporter: | pav |
Component: | gabble | Assignee: | Telepathy bugs list <telepathy-bugs> |
Status: | RESOLVED NOTOURBUG | QA Contact: | Telepathy bugs list <telepathy-bugs> |
Severity: | normal | ||
Priority: | medium | CC: | anotherjonathon |
Version: | 0.8 | ||
Hardware: | All | ||
OS: | All | ||
Whiteboard: | |||
i915 platform: | i915 features: |
Description
pav
2010-02-20 09:06:38 UTC
*** Bug 26690 has been marked as a duplicate of this bug. *** This seems likely to be related to jabber.org's switch to M-Link; I can't remember any reports of timeout/idle problems when they were still running ejabberd. (In reply to comment #0) > (telepathy-gabble:9195): gabble-DEBUG: connection_iq_unknown_cb: got unknown > iq: > <iq id="ping" type="get" to="xxx@jabber.org/Telepathy.a7f3a1a3" > from="jabber.org"> <ping xmlns="urn:xmpp:ping"></ping> > </iq> I wonder whether M-Link assumes that all clients support XEP-0199 XMPP Ping, and drops clients who don't respond successfully? Current versions of Gabble don't support this feature, but I believe we correctly reply with an error to any IQ we don't support, which in the case of a ping should be an equally useful indicator that we're still alive. If you turn on Loudmouth debug (LM_DEBUG=net), do you see Gabble sending an error reply just after the message I quoted? Here's an excrept with LM_DEBUG=net It seems to me that it at least claims that it sends a reply. ********************************************************************** RECV [302]: ----------------------------------- '<iq from='proxy.eu.jabber.org' to='pv@jabber.org/Telepathy.2927922d' type='result' id='24868818448'><query xmlns='http://jabber.org/protocol/disco#info'><identity category='proxy' type='bytestreams' name='SOCKS5 Bytestreams Service'/><feature var='http://jabber.org/protocol/bytestreams'/></query></iq>' ----------------------------------- (telepathy-gabble:12775): gabble-DEBUG: connection_iq_unknown_cb: got unknown iq: <iq id="24868818448" type="result" to="pv@jabber.org/Telepathy.2927922d" from="proxy.eu.jabber.org"> <query xmlns="http://jabber.org/protocol/disco#info"> <identity name="SOCKS5 Bytestreams Service" type="bytestreams" category="proxy"></identity> <feature var="http://jabber.org/protocol/bytestreams"></feature> </query> </iq> RECV [115]: ----------------------------------- '<iq from='jabber.org' to='pv@jabber.org/Telepathy.2927922d' type='get' id='ping'><ping xmlns='urn:xmpp:ping'/></iq>' ----------------------------------- (telepathy-gabble:12775): gabble-DEBUG: connection_iq_unknown_cb: got unknown iq: <iq id="ping" type="get" to="pv@jabber.org/Telepathy.2927922d" from="jabber.org"> <ping xmlns="urn:xmpp:ping"></ping> </iq> SEND: ----------------------------------- <iq type="error" to="jabber.org" id="ping"> <ping xmlns="urn:xmpp:ping"></ping> <error type="cancel" code="502"> <service-unavailable xmlns="urn:ietf:params:xml:ns:xmpp-stanzas"></service-unavailable> </error> </iq> ----------------------------------- RECV [1]: ----------------------------------- ' ' ----------------------------------- RECV [110]: ----------------------------------- '<stream:error><connection-timeout xmlns='urn:ietf:params:xml:ns:xmpp-streams'/></stream:error></stream:stream>' ----------------------------------- (telepathy-gabble:12775): gabble-DEBUG: connection_stream_error_cb: got stream error: <stream:error> <connection-timeout xmlns="urn:ietf:params:xml:ns:xmpp-streams"></connection-timeout> </stream:error> Freeing up IOChannel and file descriptor (telepathy-gabble:12775): gabble-DEBUG: connection_disconnected_cb: called with reason 3 ********************************************************************** Also, asking around at juser@jabber.org if they know what's going on: http://article.gmane.org/gmane.network.jabber.user/2529 Got a reply from jabber.org admins;
on Thu, 25 Feb 2010 10:35:46 +0000, Kevin Smith wrote:
> Confirmed, this is an issue with jabber.org not with Telepathy.
Thanks for investigating. We're doing the (or, at least, a) right thing and they confirm that the problem is at their end, so, NOTOURBUG. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.