Event ID 1025 with EcEntryIdFromAddr causing failed local delivery
Hello, We're running Exchange 2003 Enterprise fully patched on top of Windows Server Enterprise 2003 R2 SP2. We have one front-end server and three back-end stores. So far this issue only appears to be happening to 2 databases (of 4) on ExchSrv1 and 2 databases (of 8) on ExchSrv3. So far none of these errors have shown up on ExchSrv2 (4 databases). Here is an example message from the Application log on ExchSrv3: Event Type: Warning Event Source: MSExchangeIS Mailbox Store Event Category: General Event ID: 1025 Date: 7/5/2010 Time: 2:29:38 PM User: N/A Computer: ExchSrv3 Description: An error occurred on database "Storage Group 1\R (ExchSrv3)". Function name or description of problem: EcEntryIdFromAddr Error: 0x467 For more information, click http://www.microsoft.com/contentredirect.asp. This log entry is the same for the other stores. These messages only appear when an e-mail destine for a user in one of the the reported mailbox store has an e-mail stuck in the local queue for delivery. Every time Exchange retries local delivery and fails another one of these messages is logged. Eventually Exchange gives up and sends an NDR to the sender. Some Googling reveals this problem was fixed in Exchange 2000 with SP2. Our Exchange 2003 servers are all scratch installs and not in place upgrades from 2000 so I doubt this bug is our problem. Somewhere else suggested stopping the SMTP queue, editing the stuck message and ripping out the 'Reply-to' line and then restarting SMTP to get the messages delivered. That is a pretty good work around but not a solution. Plus if this problem were related to 'Reply-to' lines I'd expect more mailbox stores to be reporting this problem then the same 4 over and over again. At this point all I've managed to come up with is running isinteg and a offline defrag of one of the databases reporting this problem and see if it helps. I've schedueled tomorrow night to do this and I'll report back if it fixes the problem. In the mean time anyone else have any suggestions?
July 6th, 2010 12:31am

On Mon, 5 Jul 2010 21:31:24 +0000, Eric Schewe wrote: > > >Hello, > >We're running Exchange 2003 Enterprise fully patched on top of Windows Server Enterprise 2003 R2 SP2. We have one front-end server and three back-end stores. So far this issue only appears to be happening to 2 databases (of 4) on ExchSrv1 and 2 databases (of 8) on ExchSrv3. So far none of these errors have shown up on ExchSrv2 (4 databases). > >Here is an example message from the Application log on ExchSrv3: > > Event Type: Warning Event Source: MSExchangeIS Mailbox Store Event Category: General Event ID: 1025 Date: 7/5/2010 Time: 2:29:38 PM User: N/A Computer: ExchSrv3 Description: An error occurred on database "Storage Group 1\R (ExchSrv3)". Function name or description of problem: EcEntryIdFromAddr Error: 0x467 For more information, click http://www.microsoft.com/contentredirect.asp. This log entry is the same for the other stores. These messages only appear when an e-mail destine for a user in one of the the reported mailbox store has an e-mail stuck in the local queue for delivery. Every time Exchange retries local delivery and fails another one of these messages is logged. Eventually Exchange gives up and sends an NDR to the sender. > >Some Googling reveals this problem was fixed in Exchange 2000 with SP2. Our Exchange 2003 servers are all scratch installs and not in place upgrades from 2000 so I doubt this bug is our problem. Somewhere else suggested stopping the SMTP queue, editing the stuck message and ripping out the 'Reply-to' line and then restarting SMTP to get the messages delivered. That is a pretty good work around but not a solution. Plus if this problem were related to 'Reply-to' lines I'd expect more mailbox stores to be reporting this problem then the same 4 over and over again. > >At this point all I've managed to come up with is running isinteg and a offline defrag of one of the databases reporting this problem and see if it helps. I've schedueled tomorrow night to do this and I'll report back if it fixes the problem. In the mean time anyone else have any suggestions? Well, if you went as far as finding a description of what "fixes" the problem, have you tried the workaround? Does it get the message delivered? I think I'd try tht first instead of the isinteg/eseutil. That'll take hours to complete. What about the properties of the mailboxes that experience the problem? Are the proxy addresses correct? Is the "Mail" attribute identical to the primary SMTP proxy address? Does moving one of the "problem" mailboxes to a database that doesn't have messages stuck in the local delivery queue get the messages delivered? Or does the problem follow the mailbox? --- Rich Matheisen MCSE+I, Exchange MVP --- Rich Matheisen MCSE+I, Exchange MVP
Free Windows Admin Tool Kit Click here and download it now
July 6th, 2010 1:13am

I checked out the mailbox and the 'mail' attribute is the same as the users primary SMTP entry on their 'E-mail Addresses' tab. I used ADSIEdit to look up their 'mail' attribute. Was that correct? The users X400 record appears to be correct. I honestly don't know what it should look like but it's the same as everyone else's except for the 's=' and 'g=' which reflects the users first and last name. Great idea moving their mailbox. I hadn't thought about that. I'll do it after the user goes home and get them to send a test message tonight. I also hadn't tried editing the stuck messages in the SMTP queue. I'll try that as well with the next stuck message. I just tried editing a stuck e-mail and deleting the 'Reply-to' line. According to the message tracker it was delivered successfully.I'm waiting on confirmation from the user which I should get tomorrow. The user confirmed they just got the e-mail delivered to their mailbox.
July 6th, 2010 1:36am

On Mon, 5 Jul 2010 22:36:25 +0000, Eric Schewe wrote: >I checked out the mailbox and the 'mail' attribute is the same as the users primary SMTP entry on their 'E-mail Addresses' tab. I used ADSIEdit to look up their 'mail' attribute. Was that correct? You can use the Active Directory Users and Computers snap-in to look for that attribute. >The users X400 record appears to be correct. I honestly don't know what it should look like but it's the same as everyone else's except for the 's=' and 'g=' which reflects the users first and last name. I don'tt think this has anything to do with X.400 addresses. >Great idea moving their mailbox. I hadn't thought about that. I'll do it after the user goes home and get them to send a test message tonight. Based on what you said below I'm not sure that's going to help. >I also hadn't tried editing the stuck messages in the SMTP queue. I'll try that as well with the next stuck message. > >I just tried editing a stuck e-mail and deleting the 'Reply-to' line. According to the message tracker it was delivered successfully.I'm waiting on confirmation from the user which I should get tomorrow. The user confirmed they just got the e-mail delivered to their mailbox. Okay . . . now, who's sending the e-mail with the reply-to header, and what's the domain in those reply-to headers? Is is e-mail from within your own organization? Is the e-mail from addresses within your organization but sent from another e-mail system (e.g. someone using their ISP account to send mail claiming to be from your-domain.com but with a reply-to in their-isp.com?). Are you sharing a namespace with any other e-mail systems? Is the domain in the reply-to address one of the domains in a Recipient Policy? Do the reply-to addresses belong to any objects in your AD? Perhaps a mail-enabled Contact, or even a user that's not mail- or mailbox-enabled but has a SMTP address in the "mail" attribute? Do you even have any such AD objects? If you do, try removing the 'mail' attribute. It's not useful and it can cause some odd problems. Do you have mail-enabled Contacts that don't have a populated targetAddress property (the "E-Mail address" property on the "Exchange "General" tab in ADUC)? All of them should, but if they were created by a program they may not be correctly constructed. There are other instances of this error that can be Googled that are happening in Exchange 2003 SP1. If you can't find any of the situations I've asked about it may be faster to call PSS and have them diagnose the problem. Given the small number of search results I'd bet there's something going on with some misconfigured items in the AD, though. --- Rich Matheisen MCSE+I, Exchange MVP --- Rich Matheisen MCSE+I, Exchange MVP
Free Windows Admin Tool Kit Click here and download it now
July 6th, 2010 5:35am

The user with the problem that we can reproduce easily goes like this. The e-mail is being sent from a local ISP (@shaw.ca) to our work accounts (@business.ca we'll say). The problem mail is originating from outside of our systems. The e-mail is coming from Firstname.Lastname@shaw.ca to Firstname.Lastname@business.ca. The reply-to address on the e-mail is Firstname.Lastname@shaw.ca (the originator). I do not believe we are sharing our namespace with other e-mail servers. Our Exchange servers solely handle @business.ca e-mail. I moved this users mailbox and they tried sending a test e-mail to their @business.ca account and my @business.ca account. The e-mail is stuck in two Exchange store servers delivery queues. He tried the same e-mail from Shaw Webmail and it worked perfectly. We figured the problem might be the users home Outlook is some how miss-configured but his wife using her @shaw.ca account in the same Outlook on the same computer has no problem sending e-mails. Plus his e-mails get to us they just get stuck in queue and the headers of the e-mails look fine. Shouldn't Webmail fail in the same way? Interestingly I did find a contact in AD that has the users Firstname.Lastname@shaw.ca associated with it. It doesn't appear to be mail enabled. I did some more digging and it turns out another stuck e-mail from a @yahoo.ca address has the same type of contact in AD. Also a @gmail.com address. Actually so far every stuck e-mail I've encountered the sender appears to have a non-mail-enabled contact in AD. I created an identical contact for myself and my e-mails didn't get stuck. I deleted the contact for the problem users @shaw.ca account, restarted SMTP on the Exchange servers. The e-mails re-queued and were delivered. Turns out a script creates these alternate contacts for employees upon request. We're thinking it's problem solved at this point. Thank you very much for your assistance.
July 7th, 2010 1:02am

Confirmed. If a user has a contact configured with their personal e-mail AND their mail client/webmail has the 'reply-to' address set to that same personal e-mail address the messages will get stuck in local delivery. If I delete the contact and wait 10-15min the e-mail will come through. If I right click on the contact, choose 'Exchange Tasks' and 'Establish E-mail Address' the e-mails will come through. If I leave the contact as is and re-configure the mail client/webmail to not have a 'reply-to' address that is the same as the sending address the e-mail comes through.
Free Windows Admin Tool Kit Click here and download it now
July 7th, 2010 1:55am

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics