GNOME Bugzilla – Bug 265915
My companys exchange server is crashing when I save an appointment to my calendar
Last modified: 2006-06-19 05:48:33 UTC
Please fill in this template when reporting a bug, unless you know what you are doing. Description of Problem: Steps to reproduce the problem: 1. I believe this happening when I save an appointment to my exchange calendar or I update an existing appointment. The mail administrator won't let me test anymore, since it causes our exchange server to crash. 2. 3. Actual Results: Here is the event that the exchange server shows just before it Dr Watsons…. Event Type: Error Event Source: EXCDO Event Category: General Event ID: 8217 Date: 9/10/2004 Time: 2:01:23 PM User: N/A Computer: MIAHCEMAIL2 Description: The processing of the recurring appointment with the subject "Check Network Solutions for expiring domains" in mailbox: ********* failed with error code 0x8004021e. This appointment will not be visible in the web client or other non-MAPI client. For more information, click http://www.microsoft.com/contentredirect.asp. Event Type: Information Event Source: DrWatson Event Category: None Event ID: 4097 Date: 9/10/2004 Time: 2:04:18 PM User: N/A Computer: MIAHCEMAIL2 Description: The application, , generated an application error The error occurred on 09/10/2004 @ 14:04:18.125 The exception generated was c0000005 at address 77F82ABE (RtlExtendedMagicDivide) Expected Results: I expect the exchange server not to crash How often does this happen? It has happened twice. After the second time my boss put his foot down and said I had to stop using Evolution for my calendar. Additional Information:
UGH. Never heard of this happening before. If you still have a copy of that appointment, can you attach it to this bug? Was there anything at all "interesting" about it? (Like, maybe, a complicated recurring rule or something.) Also, if you're still on speaking terms with your Exchange admin [:-/], could you find out what service pack of Exchange you're running?
Created attachment 44199 [details] This is the updated appointment
Created attachment 44200 [details] This is the original appointment
Created attachment 44201 [details] This is the other mail message that crashed the exchange server.
Dan, The original appointment I was able to save to my calendar without crashing our exchange server. However, when I accepted the updated appointment the exchange server crashed. The third appointment also crashed the server. This appointment is independent of the other two appointments. Sorry, if this does not make any sense. Exchange 2000 SP3, Windows 2000 SP3 --Lynn
Received the following mail from a person who was about to confirm this: "I've read the following Bug Report http://bugzilla.gnome.org/show_bug.cgi?id=265915 and a few days ago we confirmed this bug with Microsoft. We have an Exchange 2000 Cluster system Crashing all the time with an Evolution Client accessing one specific appointment. An export of the item with outlook and import to a testsystem confirmed that it's working for all versions of Exchange 2000 available at the moment. We do not exactly know what evolution is doing wrong and / or if we have a working fix for Exchange (We test the fix: http://support.microsoft.com/?id=875550 at the moment) I'm trying to find out what exactly is wrong on the appointment to inform you what Evolution does wrong."
Hi, You may want to cross-reference this with 67330, where I have included an appointment which always crashes the exchange back end. It seems that appointments from one or two outlook/exchange users always crash the connector, but emails from other, apparently the same kind of users, don't cause crashes.
why is this marked 2.1? we should probably look at for 2.0.x too.
Any luck in getting the offending appointment? Also could you try using 2.0.4 [ you will have build from sources for now ] and see if the fix for 67330 fixes this as well.
I have a possible similar situation, thought I'd document it here. My setup: Linux 2.6.9-rc3 (Gentoo flavor), Evolution 2.0.4, Connector 2.0.2, Exchange 6.0 Build 6249 SP 3 on Win 2000 server with SP 4. I've been using Evolution constantly on my work PC for the last 6 months; I also normally use Outlook Web Access several times per week. The problem: I was at home using OWA, and had left Evolution running on my machine at work. After reading my new emails in OWA, I clicked on the calendar icon to review my schedule. Just doing that apparently locked up Exchange, at least there were references to Evolution in the Exchange Server logs right at the time of the crash. Hope this helps.
I'm using evolution 2.0.4 and connector version 2.0.4 and the exchange server here has been having similar crashes. I was in the process of trying to add a recurring all-day appointmet with two days exclusion and received an 'Unknown Error' from Evolution. After shutting down Evolution and restarting it, I was unable to authenticate with Exchange, although I confirmed my password was correct (I was able to log into OWA)... A note here: the exchange server was recently configured to not use the HTTP login method (where you get a browser prompt asking for your login and password) and instead uses a login form. It is Exchange 2000 but I don't have any more information about the server version. ... at this point, using fetchmail with IMAP was still working, but very shortly after this it stopped. Presumably, the exchange server had crashed. Now, even trying to access the calendar via OWA causes the server to crash. A coworker met with a similar fate, but his attempted appointment was simpler than mine- non-recurring. In both cases, the appointment reported as a problem by exchange was not the same appointment we tried to add, and neither of us had edited the problem appointment (in my coworker's case, the problem appointment was a recurring appointment added long ago, not using Evolution; in my case it was a recurring appointment added not-so long ago, using Evolution, but hadn't caused this problem in the past). Hope this helps.
I have just encountered the same scenario. Created a meeting in Evolution 2.2.1.1, (Ubuntu 5.04). Simple reccurrance, once per week. Sent invite, no problems were evident. When the invite was accepted by the other party, the following errors were generated in the Exchange server's applog: Event Type: Error Event Source: EXCDO Event Category: General Event ID: 8217 Date: 5/26/2005 Time: 4:30:55 PM User: N/A Computer: {servername} Description: The processing of the recurring appointment with the subject "" in mailbox: {myfullname} failed with error code 0x8004021e. This appointment will not be visible in the web client or other non-MAPI client. For more information, click http://www.microsoft.com/contentredirect.asp. Event Type: Error Event Source: EXCDO Event Category: General Event ID: 8217 Date: 5/26/2005 Time: 4:30:55 PM User: N/A Computer: {servername} Description: The processing of the recurring appointment with the subject "Weekly Update" in mailbox: {myfullname} failed with error code 0x800cce09. This appointment will not be visible in the web client or other non-MAPI client. Moments later, the Information Store service died. It restarted, and appeared normal. I then went through OWA to view the appointment (glutton for punishment!) and OWA hung, ExchangeIS crashed, and the whole service restart process began again. Ouch!
Sebastian: If its posible for you to attach ics file of appointment which is causing server crash, please attach it here that would help us to reproduce the bug. You can even mail it to surf / me. At the same time we will try at our end to reproduce the same.
From Outlook 03, I cannot export the appointment. Outlook creates an empty file, then gives a dialog "The operation failed". The appointment does show in the preview window. I can create a .msg file from Outlook. I can then open that msg file, and it shows up normall on screen. If I attempt to save this new file as .ics, it still gives the same error. I will email you the .msg file.
By the way, this is Exchange 2000 with current patches.
I had a similar problem with my company's Exchange 2000 server yesterday. Seems that whatever my Evolution client did seems to have caused the information store service to shutdown on the Exchange Server. It is Exchange 2000 with ALL of the latest service packs and hotfixes. I was saving/updating a recurring appointment that had a single updated occurance along with it for that day. I am not allowed to "test" anything as it took out the entire company's email while the information store was recycling so I don't know the exact steps to reproduce it. Do we have any more info from either the Evolution developers or from Microsoft about this apparent bug? Here's one of the NT Event logs from the server: The calendaring agent cannot bind to the appointment with the subject "Accepted: Updated: Status Meeting" in mailbox MY Mailbox because the appointment is missing a start time. The calendaring agent will not publish free/busy information for this appointment. To correct this, please delete the appointment and recreate it.
I just wanted to mention that I experienced a very similar situation. In my case, however, the exchange server crashed when I attempted to view my Calendar in OWA; Evolution wasn't even running anywhere at the time, although I had used it to update some appointments. My administrator called me up and informed me the server had gone down and that the log indicated my computer and, more importantly, Evolution as the culprit. I was asked not to use Evolution any further for company email. What a shame :( Thus, I don't believe it will be easy to obtain much information. If there is anything I can contribute, such as a log file, data store file, etc., feel free to contact me.
Ok, another "me too": in May, I had experienced the same problem, where Exchange server started crashing and it was crashed to me. I've been banned from using Connector, and doubt I will ever be able to use it at this company again. At first I thought it was impossible for it to be Connector, since the crashes were happening even when I was doing nothing in the Evolution window. But seeing this bug report, clearly it really was the cause. (Well, it's still an Exchange bug ultimately, but Connector needs to make sure it doesn't trigger it.)
Changing state of the bug to New as i remember Sebestian has mailed the appointment to Surf. Now we should try to reproduce this on our test setup.
I get the following everytime trying accept a instance of a recurring appointment with no master object whis is actually of type cdoException value:3. The request is sent thrice in a loop from evolution. The thrid response says its created although it isn't. I think this must be the case which is crashing the server. First Response: 412 Precondition Failed E2k-Debug: 0x80f8138 @ 1133884557 MS-WebStorage: 6.5.6944 Date: Tue, 06 Dec 2005 15:51:03 GMT X-Powered-By: ASP.NET Server: Microsoft-IIS/6.0 Connection: close Content-Length: 1678 Content-Type: text/html 7 Connection terminated unexpectedly E2k-Debug: 0x80f8138 @ 1133884557 ---------------------------------------------- Second response 401 Unauthorized E2k-Debug: 0x80f8138 @ 1133884557 Date: Tue, 06 Dec 2005 15:51:03 GMT X-Powered-By: ASP.NET Server: Microsoft-IIS/6.0 Content-Length: 83 Content-Type: text/html ---------------------------------------------- Third response 201 Created E2k-Debug: 0x80f8138 @ 1133884557 MS-Exchange-Permanent-URL: https://164.99.168.136/exchange/pnayak/-FlatUrlSpace-/7b3d0aac42efb943879b4414f96b57ce-29c4a/7b3d0aac42efb943879b4414f96b57ce-2b4dc MS-WebStorage: 6.5.6944 ResourceTag: <rt:7b3d0aac42efb943879b4414f96b57ce00000002b4dc7b3d0aac42efb943879b4414f96b57ce0000000521fe> GetETag: "7b3d0aac42efb943879b4414f96b57ce0000000521fe" Date: Tue, 06 Dec 2005 15:51:03 GMT Location: https://164.99.168.136/exchange/pnayak/Calendar/Updated:%20***************%20hey%20hey%20asssssssss-12.EML X-Powered-By: ASP.NET Server: Microsoft-IIS/6.0 Allow: OPTIONS, TRACE, GET, HEAD, DELETE, PUT, COPY, MOVE, PROPFIND, PROPPATCH, SEARCH, SUBSCRIBE, UNSUBSCRIBE, POLL, BDELETE, BCOPY, BMOVE, BPROPPATCH, BPROPFIND, LOCK, UNLOCK Content-Length: 125 Repl-UID: <rid:7b3d0aac42efb943879b4414f96b57ce00000002b4dc> Content-Type: text/html
This should have been fixed now. Couple of fixes have been committed to CVS head on March, 5th, 2006. Can any of you please try connector from cvs head or the 2.6 which will be released on March 15th, 2006.
Please re-open, if the issue still exists.