Jump to content

Direct Push issues...


Guest andyhud

Recommended Posts

Guest andyhud

Ok, this I admit is a long shot, and I apologise now if people feel this isnt right for this forum due to its technical aspect. I also apologise for giving what seems a long story to a situation (I'm hoping) others have come across. I'm a sucker for providing every bit of information regardless of how useful it is.

Ok, situation...

I have at home 8Mb ADSL with static IP (x5), firewall and exchange 2003 sp2 server. I've had this running now for 4 years + and its worked brilliantly. I use OWA (via SSL) I host my own DNS and to cut a long story short, it all works.

Finally last week I got my hands (via ebay) on an M3100. I had it unlocked (as I'm with T-mobile with the WnW package for

Link to comment
Share on other sites

Just to let you know what config I've got that seems to be working:

in ESM, Mobile Services - All 6 tickboxes ticked.

in IIS manager. Website with exchange VDir - Website tab -

- HTTP keep-alives enabled, timeout 120

Home directory tab - App config - Options tab

- Enable session state, ticked

- Enable buffering, ticked

Just the most prominent settings that could affect directpush listed there. If you want to know any other settings, let me know.

/J.

Know it's covering the obvious, but you have made sure you've got ur password saved for the activesync server connection on the phone?

Edited by JezB
Link to comment
Share on other sites

Guest andyhud
silly question time...have you done a manual activesync on the device?

Gents, thankyou so far

Jez. thanks for the info.. I'll check it against mine shortly...

Jim.. yes, manual activesync has always worked fine on my old MDA compact and on my M3100. Its just the directpush "feature" that is failing to work...

Can anyone check their APPLICATION event log on their exchange server for event ID 3025 (IP AUTD Initialized)... I'm not getting that on my (but I think a reboot is in order anyhow) server when IIS Admin and its dependencies start up...

Huff....

Link to comment
Share on other sites

Guest andyhud

Jez ok... checked all those settings in my IIS.. they all appear to be "defaults" in the default website and yes, mine are identical to the ones you listed...

I can only assume some major debugging is in order..

Another thought... if someone has their own exchange server they can access, can you go to a command prompt and do a "netstat -ano"

You need to look for UDP Port 2883 and see if its listed... it should be... (apparently MS reckon)

... Clutching at straws....

A.

p.s. Password is definitely saved in activesync on my M3100...

Edited by andyhud
Link to comment
Share on other sites

Guest jimbouk

1) You say that the security enforcement was passed down to the 3100. Was that over 3G/GPRS or via the USB to a pc connected to the web?

2) Should have explained myself better re manual sync. Can you now initiate an activesync over GPRS to the server ?

Link to comment
Share on other sites

Guest randomelements

As jimbo says can you do a manual wireless sync and then check your event log. The 3025 should be triggered after this.

It may be worth looking for an event 3005 "Unexpected Exchange mailbox Server error"

As you are using SSL have you been through http://support.microsoft.com/default.aspx?...kb;en-us;817379 to prevent the SSL from causing EAS problems?

Link to comment
Share on other sites

I know it's obvious but didn't see it mentioned, check that you hace activesync on the phone set to sync as new items arrive and that the schedule is valid.

Personally, I would now start by turning on verbose logging on Activesync on the phone, this will give you a good idea on what is going on.

Edited by greyt
Link to comment
Share on other sites

Guest StGeorge
So... I'm just wondering what others had done, if anything, to get there direct push to work using their own exchange servers. I appreciate that people with hosted exchange accounts can't comment on the exchange server configuration they use.

Andy

I had a real nightmare getting direct push to work on my device! My setup differs quite alot from yours as I am using SBS with Exchange and do not have a static IP. I finally managed to get direct push working but not without several hours head scratching!!

From reading your post I am sure you have checked this BUT here goes....... the problem I had was that the configuration of OMA was wrong and as a result OMA wasnt working. Once I fixed this everything fell into place. The reason I ask is that you mention specifically you have sent an email via OWA, can you do the same via OMA?

Like I said sure this is a red herring but it was what was causing my sync to fail!!

Now if only I could find a way around the pesky 15min send and receive restriction on the POP3 Connector I would be really happy.

Good luck mate

Ed

Link to comment
Share on other sites

Guest andyhud

Good Morning Gents,

Currently back at work and no change on the directpush front, however I have noted that a sync was performed 'randomly' at 7am ish this morning while I was asleep. However, nothing relating to Event 3025 was logged in my Application Event Log on the server at that time....

Anyhow... back on track, I will now try to answer your comments/questions you guys posted last night/this morning.

Jim,

1. Yes, the password enforcement was passed down to the phone via a 3G/GPRS connection. (I havent even configured USB activesync on the phone yet, I entered all the OTA (Over-the-air) sync information manually once I put the new Tytn ROM on it)...hmmm, thinking about that, perhaps I need to do a USB sync with my outlook on my pc/laptop first... but I would of thought that wouldnt of made any difference.

2. Yes, I have also been able to manually initiate a manual sync from the device (by pressing "sync" !!) on the device to the server over both 3G and GPRS. I've just done it now infact... all fine... 5 new emails.. but they didnt come down of their own accord as I want.

Randomelements,

1. I do indeed get the following error. This one was received at 20:43 last night, however I have done a manual sync since then without that error being generated.

Event ID 3005:

Unexpected Exchange mailbox Server error: Server: [sERVER.mydomain.co.uk] User: [[email protected]] HTTP status code: [501]. Verify that the Exchange mailbox Server is working correctly.

I have used www.eventid.net to try and troubleshoot this and found this article a couple of days ago: http://www.eventid.net/display.asp?eventid...ync&phase=1

I wouldnt say I have "excessive calendar entries... perhaps 1 or 2 a day... hardly what I deem excessive. What I do note however is that this error wasnt ever logged when I was manually syncing (every 5 minutes) on my old WM2003SE device.

2. KB817379... Yes, I have done this yonks ago and its been fine every since MS released the correct version without the typo errors in it. My ExchDAV settings are correct. I've double checked them.

Greyt,

1. Yes, the phone is set to "As Items arrive" on both peak and off peak. Basically set 24/7. still no joy.

2. Verbose logging is enabled and I've been told to scour through the files in \windows\activesync on the device to look for the http "ping" command that hops back and forth between the server and the device to keep it in sync. I'm struggling to find that at the moment.

StGeorge,

Indeed, OMA does play an integrale part of activesync and I can confirm my OMA does work correctly and always has.

FYI, the 15min pop3 connector is hard coded by MS and cant be changed. I asked about this over a year ago. Your only bet is to use another pop3 connector (IGETMAIL works brilliantly), or get a static IP and SMTP feed.

.... so all in all a a bit bizarre still.. I think I'm going to reboot my exchange server (albeit remotely) shortly just for the sake of argument. I'm also going to hunt for those eventlog messages further and see if they crop up on boot up etc and during a sync... It is bizarre it successfully receives the password enforcement policies fine but not directpush...

If anyone has had time to see when Event ID 3025 is logged in there exchange server application log (either just as the server boots up/IIS Admin service is restarted) or during a 'directpush' I'd be grateful.. that way I know when to keep an eye out for it...

I may end up rebuilding my IIS.

Any other thoughts/comments/suggestions are always gratefully received... I hope I'm providing enough information.

Cheers

Andy

Link to comment
Share on other sites

Guest andyhud

WEIRD UPDATE

Ok... after posting that reply I checked my server event log , and at 7:07 am this morning I did get (for the first time as far as I can see) Event 3025 - IP Based AUTD Initialized

Also, doing a netstat -ano from the command line returns showing UDP Port 2883 running which it didnt yesterday.... thing is... I havent done anything yet!

That was the only time I got it though and I did receive a lot of email during the course of the night (as I always do)...

...its as if its 'sort of working' ....

Darn....

Edited by andyhud
Link to comment
Share on other sites

There should be two activesync logs that are created, as you say these would normally show the ping and keep alive transactions etc.

If you sync the phone manually and then quickly send yourself an email does it work then? I am trying to see if push email is working but you have a timeout issue along the chain from Exchange to the phone.

There is a chance you may get some info from the IIS logs on the server, but since manual synching is working there is less chance this will be helpful

Link to comment
Share on other sites

Guest andyhud
If you sync the phone manually and then quickly send yourself an email does it work then? I am trying to see if push email is working but you have a timeout issue along the chain from Exchange to the phone.

There is a chance you may get some info from the IIS logs on the server, but since manual synching is working there is less chance this will be helpful

Greyt,

Indeed, If I sync the phone manually and send the email I get it on my phone. I've just done another now, from my work email to my home email (the activesync account) and I'm sitting here with the phone in my hands on the activesync screen begging for it to start a sync... but nothing... despite the 7am weird one...

This is doing my nutt in... I need to methodically work through where it might be failing...

I mean, does "Event ID 3025 - IP Based AUTO Initialised" get logged EVERY time the server does a directpush? or is it just once as it loads for the first time after a boot/reboot. I need to ascertain this so it doesnt send my on a wild goose chase...

Cheers

A.

Link to comment
Share on other sites

I mean, does "Event ID 3025 - IP Based AUTO Initialised" get logged EVERY time the server does a directpush? or is it just once as it loads for the first time after a boot/reboot. I need to ascertain this so it doesnt send my on a wild goose chase...

Cheers

A.

3025 is a one off entry on Exchange startup

Link to comment
Share on other sites

Guest colossusuk

I suggest starting off with the basics first.

Can you get Push email working over the default IIS configuration , NO SSL over port 80 http ?

If you can get that working then we can go from here.

Edited by colossusuk
Link to comment
Share on other sites

Guest randomelements

Andy,

The 3025 just gets triggered once to set things up and as far as I could see from my logs this happens on the first manual sync or heartbeat.

While looking at this I broke my own direct push by mucking about with the Min/MaxHeartbeatInterval keys LOL

I removed both keys and rebooted, did a manual sync and then re-added the MinHeartbeatInterval key only (as the Max is normally hard coded to 59 minutes) and it started working again. The value I used was Hex 30 (Dec 48).

Don't know if it's worth having another look at your heartbeat.

Colin.

Link to comment
Share on other sites

Guest andyhud

Ok.. further update as of midday - thursday...

I decided to reboot the server (remotely as I'm at work).. so yeah.. I did that and it came back up fine... amazingly my phone started a sync by itself once the server rebooted... I was a little suprised but didnt assume it was fixed.

So I've sent another test email to my activesync account and lo and behold.. no directpush... Its just sitting there saying "Synchronized: Today 11:55" (when it came back up from a reboot)

Indeed I have looked at the registry keys, applied them (some) , restarted the IIS Admin service and tested, but still to no avail.... I havent tried this again since the reboot just now (as in adding the reg keys) but I think it maybe something I will try....)

Colossusuk, while I agree that is a sensible approach I just dont want to go switching off the SSL to test it via port 80 when I know all other functions via SSL 443 are working fine. I admit however, that is push comes to shove, then that is what I'll do.

I'm running out of options... fast

A.

Link to comment
Share on other sites

I'm running out of options... fast

A.

I've just posted some links to documents various to do with "Direct Push" - see http://www.modaco.com/Push-Email-t245158.html for all the gory details - and I was wondering if yours might be a firewall issue?? Your original post says "blah, blah, firewall, blah, blah", but you don't elaborate any more than that (and I don't blame you for being security conscious in that respect!!). If you are using ISA2000 (some of us still do because their boss is too mean to upgrade!!!), one of the links I listed in the above post has some information about that....

Good luck in your quest to get this working!!

Cheers,

Mark.

Link to comment
Share on other sites

Guest jimbouk

If you think the server (and any proxy en route) is set up correctly, then I would hard reset the device, then set it up to sync with the exchange server via a new partnership on a pc, then after it has sync'd, disconnect it and turn on directpush and see if that sorts it.

( you could try deleting the partnership and then setting it back up without hard resetting it first).

Link to comment
Share on other sites

Guest andyhud
If you think the server (and any proxy en route) is set up correctly, then I would hard reset the device, then set it up to sync with the exchange server via a new partnership on a pc, then after it has sync'd, disconnect it and turn on directpush and see if that sorts it.

( you could try deleting the partnership and then setting it back up without hard resetting it first).

Jim, I think I may be heading along those lines to be honest.... I actually met up with 'randomelements' early today to discuss (and see his Vario II) as we both work in Canary Wharf... its bizarre to say the least and I think I'm heading that direction...

Mark... yes, you are correct in both respects. I'm actually running ISA 2004 and I have checked the MS articles relating to SSL timeouts etc but that again doesnt appear to be working.

I do have a spare ADSL router/hardware firewall, so I'm thinking of configuring that up for my internet access and testing it through that to rule out my ISA 2004 server.

Question is... what do I do first....

More food for thought...

Thanks again for everyones assistance in this.

Andy

Link to comment
Share on other sites

Guest jimbouk

Simply deleting the partnership in activesync on your pc, then reconnecting the device, in order to reset the sync relationship between the device and the exchange server is the least painful.

I admit that the whole proxy to BE server to FE server gubbins is beyond me - although I assume that this is your own exchange server and thus probably one box and thus should be simple!

Link to comment
Share on other sites

  • 2 weeks later...

Andy

I read you are using ISA 2004 with OWA over SSL. Are you using FBA? If so, have you configured FBA on at the ISA, or on the front-end (or just mailbox server if no front/back)?

The only way to have OWA FBA on ISA 2004 and have OMA and RPC/HTTP and EAS work is by frigging the HTTP virtuals, ie there's two. This is a common config for those with 1 NIC/IP on the ISA, explained in this article http://support.microsoft.com/kb/817379/ (its plausable you have implemented this)

Others configure, FBA at the Exchange server and set the web listener on ISA to Basic (obviously using SSL).

Basically, direct push talks to the first virtual using webdav subscribe, and hence fails. More here http://support.microsoft.com/?kbid=916640

Either way, give us a shout on how you are getting on.

Timiano

Link to comment
Share on other sites

S'okay - got a mail from Andy last week saying that it's now working.

Chaps,

Indeed all is working well...

Timiano, I am using FBA / ISA 2004 and all articles had been followed and adhered too..

In the end I just deleted all the IIS folders (backed up database first) and then reset them all again and ran 817379 once more and rebooted..

All started working after that...

Cheers for everyones help.

A.

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.