Well, in one of yours posts it says that your foldersync takes 220 seconds. That’s way too long, especially as Androids have a hard timeout of 30 seconds (for the request to end).

Improving speed between components is definetely the best way to deal with that. You can also enable the USE_PARTIAL_FOLDERSYNC feature in your config that actively terminates your requests if they would take too long.

Good that it’s solved.

Cheers,
Sebastian

Well, you need to understand that what you want/like is not always possible. This spec is written by Microsoft and we have to follow the spec for compatibility.
They surely believed “no one will ever do this”, but we know, technicians like to be proved wrong by users… ;)

Please also understand that several people of these 820 recipients could not see your email because their email clients could/will also not handle this.
Insisting in doing it might be “OK for you”, but be aware that people could actually not get your message.

What I mean is, just understand this is not a bug but your choice.

Cheers,
Sebastian

Thanks for the investigation and the fix. I applied the changes to our installation and it works :-)

I had this issue also without the Teamviewer extension (just checked, it’s installed but deactivated).
Latest KOE 1.3 has fixes for this behaviour and the calendaring is working fine now.

There is still a similar issue that happens when flagging emails after moving them.
https://jira.kopano.io/browse/KOE-103
We haven’t found a solution for this yet.

Cheers,
Sebastian

The issue is fixed now, root cause identified.
The new fix is currently in develop and will be merged into release/2.3 soon to go into 2.3.7.

More information is in ticket ZP-1199.

I just tried to reproduce this case with our Mobile Lumia 640 with Windows Phone 8.1 Update 2.
It just works. The contact is removed from the server but neither other contacts or the contacts folder are compromized.

Thanks @Sebastian - nice and clear answer,

its not that importent, because webapp still got this remionder and if the user want to have it in outlook too, he can just edit the calendar item and set a reminder - This reminder keeps popping up how it should be :)

Coffee_is_life

The loop detection is somewhat normal, depending on the circumstances. It could be that some appointments or contacts are also broken. Detailed information about how it works is available here: https://wiki.z-hub.io/display/ZP/Loop+detection

You could run clearloop with z-push-admin but depending on the issues z-push finds it will just start loop detection again.

Can you run kopano-fsck on these stores?

Cheers,
Sebastian

I doubt that NC/ OC itself causes the issue. The Univention (UCS) NC package installs NC within a docker container. I am guessing, that there is more automatically installed by the UCS app center than only docker and the corresponding NC image. Especially the fact, that UCS runs NC within a docker container let me doubt, that NC/ OC is the reason.

I posted this case already into the univention forum, but got no feedback at all already. Maybe everybody is at the CeBIT at the moment.

Hi Bernd-G,

I’ve just tested it on my KC 8.2.1-530 with Outlook 2013, Webapp 3.2.0.335-19.1 and Z-Push 2.3.5 and the message body was there in webapp for the attendee.

Manfred

Yeah, quite possible that ZP-1158 influenced in this.
The bug fixed in ZP-1158 normally only happens if you have emails that are not 100% MAPI conform, e.g. they were imported from somewhere else.
They cause OL to stall because of one flag OL is not able to handle.

Either of one could have fixed it. The shorter sync interval could mean that the(se) faulty email(s) were never sent to OL -> no issue, or the fix of ZP-1158 fixed them so that OL can process them correctly.

Anyway, good that it’s working now.

Cheers,
Sebastian

Hi Thomas,

when you unhook and rehook a store the internal ids of the users change. What happened in the case above, that the internal references to that user and it’s emails address get lost during this process.
When synching z-push tries to solve that reference to the GAB but does not find the user.
From the Z-Push side there is no way to fix this (not sure if it’s fixable at all).

We can fix the warning of your log. What happens there is very similar: the link to the attendee of a meeting got lost and we can’t determine the attendee status anymore.
We can fix the warning, but not the underlying cause of the issue. https://jira.z-hub.io/browse/ZP-1160

Perhaps Kopano support can help with that.

Cheers,
Sebastian

Hi Felix,

yes, I am using that ebuild - because there is no official gentoo package so far.
As gentoo is a meta distribution, it takes the actual repoistory from git and compiles it on my gentoo machine.

I just tried and deleted my profile in Outlook and recreated it.
Device ID canged when having a look with z-push-admin - but the error is still present, when I delete an email.

Edit:
the log entry does not appear, when I move/delete a single mail.
it appear even after creating a new profile, when deleting all mails via “right-click” --> “Ordner leeren” (means: delete all mails in trash)

My Outlook computer is not in the same network as the server. It is connected via DSL and VPN.
Maybe the problem is just a too slow connection that leads to sync problems.

Greetings
Vogi