Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck)
-
Hi @pax,
FYI - after the downgrade - it did not work immediately - it took one day…please don’t ask why. I also initialized a resync at the koe and at the z-push admin…
Good Luck to fix it! -
@anotherandy said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
it did not work immediately - it took one day
what the hell…maybe Outlook and Z-Push have to drink a cup of coffee and eat a big piece of cake to befriend each other
I will give them a weekend and a server restart
-
Hello,
we have the same problems. In the meantime, the Outlook profile of one account was completely deleted, including the OST files, and Outlook was brought up to date with a complete online repair installation and only then started with a completely new profile.
This did not help either.
Unfortunately, there is nothing helpful in the logs.
This is Kopano4UCS:ii kopano4ucs-z-push 1.5.33 all Meta package for Z-Push installation ii z-push-backend-kopano 2.6.1+0-0 all Z-Push Kopano backend ii z-push-common 2.6.1+0-0 all open source implementation of the ActiveSync protocol ii z-push-config-apache 2.6.1+0-0 all Z-Push apache configuration ii z-push-ipc-sharedmemory 2.6.1+0-0 all Z-Push ipc shared memory provider ii z-push-kopano 2.6.1+0-0 all Z-Push for Kopano ii z-push-kopano-gabsync 2.6.1+0-0 all GAB sync for Kopano
Outlook hangs in “connected” status with “calendar being updated”.
All conflicts and potentially problematic contacts, appointments and tasks have been resolved. Still no synchronisation.
We are now at our wit’s end…
Best regards,
HP -
hi @HPH_ZForum,
that sounds bad…I suggest to wait what the results @Pax “Teatime” and Reboot are.
My Outlook works again very well - if Pax’s Outlooks are starting to work again the issue might be z-push.
Meanwile you can start checking if UCS allows package downgrades…
br
Andreas -
Hi @Pax and @HPH_ZForum,
news! - my sync was stuck again!
Yesterday is sent an event with “ü” in the subject - and due to an z-push error I started a full resync via KOE…and it got stuck.I updated the event to “UE” and I am now trying a resync…
If it does not work - I just remembered this:
https://forum.kopano.io/topic/3577/z-push-apache-german-umlaute-öäü-are-not-displayed-correctly/3?_=1610883701707
My next try is to change the language in
vi /usr/share/z-push/backend/kopano/kopano.php
I will keep you up 2 date.br
Andreas -
Update:
I found another event with ü in the subject and removed it.
Also I changed the /usr/share/z-push/backend/kopano/kopano.php to german uft8.However, with the changed kopano.php and no event with ü or ä or ö my outlook stated syncing again, until the calendar.
Here it is stuck atSynchronization progress:
Folder: Kalender Sync: Synchronizing Status: 4% (83/2333)
and the sync was damn slow…my z-push.log shoes this entries while he tries to sync the calendar:
17/01/2021 14:27:13 [29419] [WARN] [andreas] KopanoBackend->GetKoeGabBackendFolderId() Found 0 entries in the store ‘SYSTEM’ matching the name ‘Z-Push-KOE-GAB’.
17/01/2021 14:27:13 [29419] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f)
the same i get in my z-push-error.log
17/01/2021 14:16:22 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:22 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:22 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:22 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:23 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:23 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:24 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:24 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:24 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)
17/01/2021 14:16:24 [10766] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f) (2)my kopano-server.log shows:
Sun Jan 17 14:16:22 2021: [warning] K-1515: Object not found unknown user “andreas@xxx.de”: andreas@xxx.de not found in LDAPBut my openldap server is up and running…authentication works without any problems…
I have now done this, according to
https://forum.kopano.io/topic/848/warning-k-1515-object-not-found-unknown-user/2
kopano-admin --syncand I checked my slapd… here I get some warnings:
Jan 17 13:02:58 xxx slapd[1444]: <= mdb_substring_candidates: (sn) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (mail) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (givenName) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (sn) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (mail) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (givenName) not indexed
Jan 17 13:03:30 xxx slapd[1444]: <= mdb_substring_candidates: (sn) not indexed
Jan 17 14:24:44 xxx slapd[1444]: <= mdb_substring_candidates: (cn) not indexed
Jan 17 14:24:44 xxx slapd[1444]: <= mdb_substring_candidates: (uid) not indexed
Jan 17 14:24:44 xxx slapd[1444]: <= mdb_substring_candidates: (mail) not indexednot sure if here is anything to do…
so far - I will give you an update if my calendar finishes the sync or not…and If I have to fix the indexes in my openldap which probably is my next doing
br
Andreas -
My last update for today - the calendar sync succeeded after 2 hours with one error,
Broken object: ‘SyncAppointment’ ignored on ‘2021-01-17 16:56’
Information: Subject: ‘XXXX’ - On: ‘2021-02-12 17:00’
Reason: Message was causing loop (2)
Item/Parent id: U8ea05:6e04b361dd504814a09c06c42378d470fc2402000000/U8ea05I now deleted this appointment and everything is fine, my Outlook is back in Sync…
My last action for today was to perform again a full sync,
started on 17:20 finished at 17:39 - with about 22K items.
Performance as expected and the calendar sync was very quick this time with less then 5 minutesThere were still the same kind of errors in the server.log and in der z-push.log…
maybe @Manfred can you give me some advice how serious these errors are?br
Andreas -
@anotherandy said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
Item/Parent id: U8ea05:6e04b361dd504814a09c06c42378d470fc2402000000/U8ea05
I now deleted this appointment and everything is fine, my Outlook is back in Sync…Is it possible from the server side to delete a corrupting calendar or email item ?
My users have outlook 2016, although i’m still trying to crack the case, so far it seams someone sending a reply to a meeting item (which i was not able to see) has caused problems in syncing for a few users. (my problem some of my users work oversea, and dont follow instructions that well, so to me its hard to get some debug info.)
-
@handyman said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
Is it possible from the server side to delete a corrupting calendar or email item ?
Within Webapp you can clear the whole calendar.
Change to calendar -> change view to “LIST” (STRG + ALT +5) -> mark everything and delete, or move them to a backup calendar somewhere else.
I did it with one user and Outlook startet syncing right after.@anotherandy said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
I suggest to wait what the results @Pax “Teatime” and Reboot are.
guess what…didnt work :D
Im trying with older Outlook versions now.
@everyone please note
Outlook is still syncing Mails even stuck at the calendar sync BUT the syncronization is so horribly slow so that there are strange things happening. Freshly sent items arent under sent items. Reply-To will delete the Mail from the inbox, until synced again some minutes or hours later. Recieved invitations arent displayed and so on. So please dont trust Outlook to show you the truth about your Mails. -
no luck with multiple older releases. Switch between x86 and x64 made no difference either.
My last try was an very old Outlook 2016 Pro 16.0.5056.1000 x86 on a Windows Server 2016 which itself recieved its last Windows Update on 15.10.2020. So this surely worked bevor.My new guess is, something within the last update to Z-Push 2.6.1 or Kopano killed something which isnt reverted if you downgrade Z-Push.
OR
the new Outlook versions are creating new type of calendar entries / invitations which gives Z-Push the creeps. -
@pax did you see the reply in https://forum.kopano.io/post/19174?
The mentioned fix is https://stash.z-hub.io/projects/ZP/repos/z-push/pull-requests/873/overview (as this cannot be seen in Jira without being logged in)
-
@fbartels said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
@pax did you see the reply in https://forum.kopano.io/post/19174?
The mentioned fix is https://stash.z-hub.io/projects/ZP/repos/z-push/pull-requests/873/overview (as this cannot be seen in Jira without being logged in)
Hey Felix i’ve ssen your post but unfortunately reverting back to Z-Push 2.5.2 doesnt make a difference.
Is there a way to revert, edit, or correct the “mapiprovider.php” mentioned? Do you know where it is stored under Ubuntu? -
@pax did you also trigger a resync? I am not really involved in the topic, but had the feeling Manfreds reply got under as in my eyes it lists what needs to be done.
The file is by default located in
/usr/share/z-push/backend/kopano
. -
Hi there,
We’re also having issues with stucked calendar syncs (Kopano Stable/supported, just downgraded from Z-Push 2.6.1 to 2.5.2, Outlook 2013/2016 on premise).
What has to be changed in mapiprovider.php to fix that issue ?Thanks
Peter -
--------Solution: --------
a simple edit of z-psuh mapiprovider
Ubuntu users will find the mapiprovider.php in /usr/share/z-push/backend/kopano as mentioned by Felix.make a backup of that file
cp /usr/share/z-push/backend/kopano/mapiprovider.php /usr/share/z-push/backend/kopano/bak-mapiprovider.php
Edit the mapiprovider.php.
Look for:if (isset($row[PR_RECIPIENT_TYPE])) { $attendee->attendeetype = $row[PR_RECIPIENT_TYPE];
change it to:
if (isset($row[PR_RECIPIENT_TYPE])) { $attendee->attendeetype = $row[PR_RECIPIENT_TYPE] != 0 ? $row[PR_RECIPIENT_TYPE] : MAPI_TO;
simply close and restart Outlook afterwards (a full resync was not necessary but hey it doenst do any harm either)
voila its working now (tested with Z-Push 2.5.2 and 2.6.1)
FUNFACT: I made this change with my downgraded Z-Push 2.5.2 after that I installed the updated 2.6.1 with apt-get install. The update reverted my edit on the mapiprovider.php and I had to redo them. So be aware of that.
@fbartels @Manfred
Sorry this was a big misunderstanding. I hadnt a Jira account so I didnt see the changes of that file. And I didnt know where to find it either.
But for you information. Reverting to Z-Push 2.5.2 and doing a full resync of Outlook isnt the solution. You have to edit the mapiprovider.php as mentioned. -
Hi @pax ,
I also had no jira account and didn’t see the fix :-( - thanks so far, I am not sure what I will do now, because currently its working…
br
Andreas -
Hello,
many thanks for the fix!
Other possible solutions, such as deleting items that were hanging in loop detection, deleting possibly problematic contacts or appointments via WebApp, setting up Outlook with a new profile again, resync etc. did not lead to anything.
After changing the PHP file, however, we actually had to trigger a complete resync via KOE, because otherwise the connection indicator in Outlook showed connection problems.
A permanent fix will also be available for Kopano4UCS with the directly integrated Kopano repos, I assume?
BR,
HP -
Hi @anotherandy said in Z-Push 2.6.1 - Outlook 365 not synching any more ( stuck):
my z-push.log shoes this entries while he tries to sync the calendar:
17/01/2021 14:27:13 [29419] [WARN] [andreas] /usr/share/z-push/backend/kopano/mapiprovider.php:274 mapi_zarafa_getuser_by_name(): Unable to resolve user: not found (8004010f)
…
my kopano-server.log shows:
Sun Jan 17 14:16:22 2021: [warning] K-1515: Object not found unknown user “andreas@xxx.de”: andreas@xxx.de not found in LDAPI guess it happens because your login name is “andreas” while mapi_zarafa_getuser_by_name tries to resolve “andreas@xxx.de”. From Z-Push point of view that’s not a big issue as there’s a fallback code to get the user. I guess the warning in server.log was also caused by mapi_zarafa_getuser_by_name.
Manfred
-
Hi all,
it looks like even with Z-Push 2.5.2 the fix from https://jira.z-hub.io/browse/ZP-1603 is required (which makes sense as there already might be calendar items in the store for which this is required), so we’ll discuss the possibility to push this change into the final repository and provide the hot-fixed 2.6.1 version so that the downgrade is not necessary.
Manfred
-
thx @manfred you’re right, my login name is andreas.
Can I fix this setting that he does not search for the email but for the username?Info @ all - if you make an upgrade from z-push 2.5 to 2.6 the fix need to be done again!
The update overwrites the
/usr/share/z-push/backend/kopano# vi mapiprovider.php
br
Andreas