It seems there’s a neverending issue with this configuration, and I’m unable to find the solution all over internet.
We have here : Spark 2.7.0 & Openfire 3.10.0 with LDAP (on Windows Server 2008 R2).
We had Spark 2.6.3 & Openfire 3.9.3 (and 3.8) with LDAP (on Ubutu Server)
In every configuration, I have the same result :
1 - When User A connects first, he sees User B not connected (it’s normal).
2- User B connects, User A doesn’t see the ‘online’ status of User B (he’s still not connected, regarding the list in the spark main window)
3- User A open a conversation with User B. Status in the conversation windows reflects the truth : User B is connected. But the main windows still displays ‘off line’ status.
Is there a solution to have real-time refresh for connection status on the main windows of Spark ?
It’s working well with Gajim for example, but not with Spark.
Thanks in advance for your kind help, and sorry for my poor english. Not my native tongue.
I notice this sometimes too but I can’t reproduce it consistently. I tend to notice it when most when the connection drops and is reestablished, and restarting spark resolves the issue. Spark dev is kind of slow right now.
Too bad Spark is underrated ! We’re using it since we dropped off Jabbin, 4 years ago.
Is there a better client than Spark to use with Openfire ? (in enterprise/company use I mean, no need of video/audio, just IM and screenshot/file sharing.)
For issue reproducing, we’ve just tested 4 servers/clients configuration these past 2 days, and we’ve 100% reproduced issue !
I saw elsewhere in this forum there’s a workaround : commenting out or delete this parameter : provider.group.className.
But I don’t know how it corrects the issue and what are the side effects. (currently trying now).
I agree. Spark is underrated, and I love how well it compliments openfire, being able to use SSO is a must for me, and one of the main reason I use it. You can try the latest nightly build to see if that helps. Also, what are you using for your backend database? are you using shared roster groups?
Nighly builds = tried and fail !
Database : embedded
Shared roster groups = no I don’t think so, we’re using Active Directory Groups in Openfire, and after small setting, we see them in Spark.
The solution with provider.group.classname is a fail too…
I think we’ll let Openfire and spark running as they run now, and we’ll see if there’s some changes in few months. (with newer version of Openfire, may be)
We’re now in Spark 2.7.2 and Openfire 4.0.1. Unfortunately, we still encounter the same “issue” : no refresh for the status of our roster in the Spark’s main display.
However, the user status is correct once a discussion window is opened.
May be I have to look after a Spark issue, instead of Openfire, no ?
whats steps are you taking to reproduce this issue so that I may try it on my end.
What version of java is Spark using?
Spark tells me this is 1.8.0_60 java version.
Lastly, we’ve installed openfire 3.9.2 in a fresh new VM in our hyper-v cluster. Then, we set it with Active Directory relationship.
After that, we installed Spark 2.7 on each PC client.
No computer is able to have a true refresh. However, “End the session” in Spark grants each people to have the current status (after restart).
As we tried 2 times, I would say this is 100% reproductible…
I"m running openfire 4.0.1, with AD, and spark at 2.7.x, without any of the issues. How many users do you have in AD, are you using shared roster groups, or are users adding contacts manually? What database are you using for openfire?
Still in embedded database.
However, regarding shared roster groups, I don’t see what it is. I’m using groups in AD and display them in Spark (some of them, at least).
you might try using something other than the embedded db, although I can’t see how that would be an issue. I’m honestly out of ideas. might be a caching issue? How large is your AD. do you have have 1000 objects that is being queried by Openfire?
There’s only few users/objects in our AD.
And regarding Openfire, the refresh status is ok in Openfire management windows, and also in Spark in the conversation window. The only issue is in the main roster window. So, I don’t think changing database will correct it. Also, I remember I tried with another jabber client (like Gajim) and it worked well…
you may want to try out the latest nightly build to see if that yields any different results. I’m out of ideas. Te presence on my main window works like a charm on all the workstations I have spark installed on. Ignite Realtime: Spark Nightly Builds