Just ran into this recently, it appears that the status being sent out by OctroTalk blows up Spark when recieved… any resolutions etc…?
I can confirm this as well (I think). Suddently this morning, Spark clients began connecting/disconnecting quite rapidly. I found that a user had recently setup OctroTalk to use with our server. I went into Server to Server Settings, blocked *.octro.net, and manually disconnected their server sessions. All seems to be well now.
I attached a graph to show the spike in packets and the subsequent drop after Octro.net was blocked.
Don’t know specifically how that applies, but nonetheless, it seems to be an issue of OctroTalk and Spark duking it out… And OctroTalk seems to be beating Spark pretty badly
Hm. This could be a Smack problem (i remember some disconnect on malformed packets issues reported in Smack forums), or even an Openfire problem. Hard to tell without deep investigation.
all i know is that when connected with a pandion client and a spark client the spark client goes into an endless reconnect cycle when somene in the roster is logged in with octro mobile
Yes, this is exactly what happened to me this morning. Once I blocked *.octro.net in Server To Server Settings, the Spark client was able to maintain its connection (and has all day since).
I’ll look into that and see if that helps at all. Don’t have someone with octro on a smart phone at my disposal at the moment.
nope, that doesnt work
An oddity that we just found out yesterday, Someone logged in with octrotalk on their phone (as we were diving to stop them) and nobody got booted…
Since the first post we have only really put in a few plugins (broadcast, fastpath, and subscription), wondering if one of those fixed it or if we have another hidden issue.
Poor Spark! People should rite better code.
Both Spark and OctroTalk must check their codes.