Scalix Forums

Hello,

we have one customer which has problems with one iPhone. From time to time the calendar and the contacts are empty. Email is running fine. A few hours or sometimes a few days later everything is ok.

The following entries appear in the log:



Do you have any ideas?

Best regards,

Alex

Now, we have the same problem with an other customer. He has no contacts on the Iphone. The stacktrace is:

The version on both servers is 11.4.5.4953

Any help would be great

Best regards,

Alex

Just had an issue with an iPhone user's contacts disappearing yesterday. Deleting the activesync account on the phone and recreating (or so they say, in a differently location) didn't seem to fix the issue. I ended up doing a sxmboxexp of the user's contacts folder, deleting all contacts from webmail, and then using sxmboximp to reimport. Contacts synced to the phone almost immediately.

Hi billb3,

thank you for your reply. I will try this. But because this is not the first time we have this problem (different customers and different users) I think this is a general issue and maybe someone from scalix can have a look on this.

Best regards,

Alex

You will either need to open a support case, or submit something to bugzilla if you want them to look...

Just a 'metoo' entry. We are also seeing this issue with an iPhone on 3.1 (?) with Scalix 11.4.5 and AS 1.0.1. Are the other iPhones with this issue on the same version or earlier?

Will try the folder recreate tasks and see if that improves the situation.

What the user is seeing is that the Contacts and Calendar items are there one minute gone the next and continue to come and go with no apparent reason or timeframe, that they can point to.

Interestingly the log I see is slightly different:

i do not know if i want to open a support ticket for this if it is obvious that we are talking about a general problem. i would prefer if anybody could tell me how we can assist scalix developers in tracing down the problem

in my case today i got a phonecall from customer that his calendar disappeared after having worked fine a few weeks now with his brand new iphone

a look in the log reveals this:



how on earth can i identify which calendar entry is suddenly causing this error?

We still have this stacktrace:


After this the contacts and sometimes also the calendar on the ipone is empty. Email is still working fine.

We don't have any support points for this customers, so I cant open an ticket.

However, it would be great if someone from scalix could at least tell us, what this error means. It is maybe a network connection problem? But we have this on two different servers with different internet connections and also over WLAN.

Best regards,

Alex

it is more useful to open a support ticket. in that case our support folks can collect full logs and maybe even get a dump of your calendar and we can analyze the issue.

A good number of things "like this", i.e. data-driven sync problems, have been identified over the last couple months and resulted in changes that we'll publish in the upcoming (due probably some time next month) 11.4.6 Scalix and 1.0.2 ActiveSync Maintenance releases.

Florian

florian wrote:it is more useful to open a support ticket. in that case our support folks can collect full logs and maybe even get a dump of your calendar and we can analyze the issue.

Hi Florian,

thank you for this advice, we have already done this (#175501), but our customer has no more support points. The problem is, that I cant tell him, that he has to buy a support packet for nearly 1000 EUR to report this problem which is maybe a bug in the activesync - component. We had enough trouble with this customer because of the deferral of the active - sync.

Please don't misunderstand me, I think scalix is a great software, but its just the manager of our customer who is affected by this problems. And he tells me constantly that his business partner (same holding, but different locations) with ms - exchange do not have any problems. So it would be great to solve this problem because I get nearly each day a call because of this.

Best regards,

Alex

Hi,

I followed the advice on some of the posts about this type of issue where it was suggested to recreate the folder. I thought I would take one step back and recreate the messages first then if that didn't work then recreate the folder and messages.

First I did an sxmboxlist of the users account
sxmboxlist -u username -l /tmp/folderlist

I checked /tmp/folderlist for the Calendar folder F-000017365, then I sxmboxexp'ed the folder to a file with a list of the message ids to another file:
sxmboxexp -u username -a /tmp/fldCALaf -f F-000017365 -l /tmp/msglistaf -v msg -F

With omcontain I emptied out the Calendar folder using option 'md' .... ONLY USE OMCONTAIN IF YOU ARE SURE YOU KNOW WHAT YOU ARE DOING AS YOU ARE PLAYING WITH YOUR USERS DATA AND ANY FALSE MOVE MEANS...YOU'VE DELETED SOMETHING AND RECOVERY RELIES ON A GOOD BACKUP! Preferably an sxmboxexp backup, a successful one at that, one you KNOW will restore okay, there is no going back once you have deleted something with omcontain!!! You have been warned.

omcontain

Option?u
User:sxadmin

User Name = sxadmin /lavender
User Number = 112
Ctr Handle = 3001
Done

Option?o
Ctr Handle (3001):
Entry Number:4 <<<<<< I knew that item 4 was the Filing Cabinet where the Calender folder resides otherwise
Ctr Handle = 3002 <<<<<< do a List first then use the number for the Filing Cabinet.
Done

Option?l
Ctr Handle (3002):
1) WASTE BASKET (FOLDER)
2) Search Area (FOLDER)
3) IMAP #Scalix namespace (FOLDER)
4) Calendar (FOLDER) <<<<<<<<<<
5) Contacts (FOLDER)
6) Drafts (FOLDER)
7) Notes (FOLDER)
Tasks (FOLDER)
9) Scalix Recovered Items (FOLDER)
10) Test1 (FOLDER)
11) MAPI Root Folders (FOLDER)
12) Journal (FOLDER)
13) New Folder 2 (FOLDER)
Done

Option?o
Ctr Handle (3002):
Entry Number:4
Ctr Handle = 3003
Done

Option?l
Ctr Handle (3003):
this will list all of the items in the calendar so you know how many you have to delete. If you don't want to delete them you could moVe them between two containers but there isn't a mv command like there is an md command.

Option?md
Ctr Handle (3003):
Enter from record number. Entry Number:1
Enter to record number. Entry Number:100

This will delete for ever 1-100 messages, you can do 1 to whatever number is the last number but I tend to do it in chunks.

Then I restored the folder with sxmboximp - which of course takes a lot longer than the export. I made sure the user was not accessing his mailbox during this whole process and keep checking that he hadn't forgotten and logged in accidently - I just know users will do this to you.
sxmboximp -u username -a /tmp/fldCALaf -f F-000017365 -l /tmp/msglistaf-new -v msg

I got a list of the messages as they were restored to compare against the list I got when the messages were stored. Not essential to get the list of messages but I was working on an MD's account, I was very cautious.

He called back once I'd told him it was done and he was able to now synch his calendar. That was yesterday and I hope I'm not putting the cart before the horse by assuming this is now fixed... S-)!

as7805 wrote:thank you for this advice, we have already done this (#175501), but our customer has no more support points. The problem is, that I cant tell him, that he has to buy a support packet for nearly 1000 EUR to report this problem which is maybe a bug in the activesync - component. We had enough trouble with this customer because of the deferral of the active - sync.

Well... first of all, I think if this is a corporate email system, there's always the chance that things can go sideways and one needs help - after all, this is typically mission critical. Therefore, I'd assume that most customers would want to have a support pack or entitlement ready, in case they need it.

Of course, if then it is determined that the problem was caused by an issue or bug inside our software, the incident (or support points) is credited back to the customers account and can be used for other things; we don't let you pay to report our bugs, obviously.

The issue is... this type of stuff is most often caused by "special" types of data, possibly coming in from external or through some mechanism. such data needs to be analyzed. you can post the data here in the forum, most people are unwilling to do that for privacy reasons, while they are happy to share it under the support contract's NDA. Also, we obviously cannot provide services here under SLA (and at the same time possibly cannibalize our service offerings), we are a commercial company, have bills to pay, too, and I'm typing this in my spare time, as we speak. I cannot personally offer to analyze that data and case for free, too.

BIt of a catch-22, then.

I guess you'll probably either have to isolate it and post stuff in limited ways, or report a reproducible bug scenario (or attach problematic sample data) in our bugtracking system, or find some way to get into that support queue, after all. Or hope that it's one of the issue we already found and fixed for AS 1.0.2 or Scalix 11.4.6.

Hm.
Florian

I would like to use omcontain to clear the calendar folder for the iPhone user on my server having a similar problem. I am unclear on how to use omcontain. I've tried to run sudo omcontain enter the password and then I get a message telling me to hit enter if I'd like to exit and then nothing happens. I don't get an interactive console that I can use to remove the content for the user.

Is there any kind of guide anywhere with information on using omcontain? I have backups of the users mailbox and calendar folder so I'm not that concerned with messing anything up. I also checked the admin guide for information about omcontain and couldn't find anything.

Thanks in advance.

just search this forum, a description on how to work omcontain is contained in many posts around here.

Florian

So I've been slogging through support steps with support and I've basically figured out the issue on my own. Scalix chokes on the data in the user's calendar and doesn't sync. The support person I am working with commented that there are many known issues with data not being processed correctly. Why haven't any fixes been released for these errors? I have two devices using ActiveSync and they both have these problems with data. I also know that when the server chokes on the data it throws a null pointer exception without any information about what actual data (a message ID would be helpful) is causing the problem. Is there anyway to get any more specific debugging information so I can figure out what specific piece of data is causing the problem?

If I can figure out what calendar items, or what types of calendar items are causing Scalix problems I can go in and delete them and solve the problem. Otherwise, I'll guess I'll just tell the user they can't use ActiveSync to sync their calendar after I sold his company on the benefits of using ActiveSync to sync their calendars with Outlook wirelessly.