Outlook Express and IMAP problems (0x800CCC0F)
Gerard Bok
Situation: 10 mailboxes on a Courier IMAP server.
Multiple users on multiple locations, all sharing all mailboxes.
The aim is to use Outlook 2003.
Couldn't get that operational at all, so for the time being, we
settled for Outlook Express instead.
Synchronize accounts produces loads of 0x800CCC0F errors.
More specific: at quiet times, I am able to sync each mailbox
individually without any error.
Once I try to sync all accounts at once, loads of errors are
reported, on random mailboxes and even when I just finished an
individual sync.
I can access all mailboxes using Pegasus without any error.
(No, switching to Pegasus is not an option. Sorry!)
And no, even with Norton Internet Security switched off, the
problems persist :-)
Questions:
- Is this setup a very unusual one ?
- Is there someone out there, running a similar setup without
problems ?
- Any hints on the cause of these problems ?
Somehow all problems seem linked to Outlook / OE. But it is
rather unlikely that all users run exactly the same versions.
Thanks!
--
Kind regards,
Gerard Bok
Jousset Maxime
"Gerard Bok" <bok...@zonnet.nl> a écrit dans le message de news:
440ec111...@News.Individual.NET...
Jim Pickering
incoming email. Disabling the option does not leave you unprotected and it
should be disabled. If that does not solve the problem then review the info
in this KB article for some other ideas:
813514 - OL: Error 0x800ccc0d or 0x800ccc0f When Receiving and Sending
E-Mail:
http://support.microsoft.com/?kbid=813514
--
Jim Pickering, MVP, Outlook Express
https://mvp.support.microsoft.com/profile=F9F51EF1-4AE3-4D23-B2D8-1171988A62D6
Please deliver feedback to the newsgroup, so that others can be helped.
Thanks.
"Gerard Bok" <bok...@zonnet.nl> wrote in message
news:440ec111...@News.Individual.NET...
Sam
> Looking for some guidance. Or aspirin :-)
>
> Situation: 10 mailboxes on a Courier IMAP server.
> Multiple users on multiple locations, all sharing all mailboxes.
>
> The aim is to use Outlook 2003.
> Couldn't get that operational at all, so for the time being, we
> settled for Outlook Express instead.
>
> Synchronize accounts produces loads of 0x800CCC0F errors.
> More specific: at quiet times, I am able to sync each mailbox
> individually without any error.
> Once I try to sync all accounts at once, loads of errors are
> reported, on random mailboxes and even when I just finished an
> individual sync.
By mailboxes I presume you are referring to IMAP folders. I doubt that you
are logging in to ten different IMAP accounts.
The default Courier-IMAP server configuration permits a maximum of five
simultaneous connections from the same IP address.
It appears that your mail client opens a separate connection for each IMAP
folder. The server refuses to accept any more connections, after the first
five, and then your mail client vomits all over itself.
Either increase the maximum number of allowed simultaneous connections, in
the server's configuration file, or switch to a different mail client that
implements IMAP properly.
Jim Pickering
key, you may have missed it):
By mailboxes I presume you are referring to IMAP folders. I doubt that you
are logging in to ten different IMAP accounts.
The default Courier-IMAP server configuration permits a maximum of five
simultaneous connections from the same IP address.
It appears that your mail client opens a separate connection for each IMAP
folder. The server refuses to accept any more connections, after the first
five, and then your mail client vomits all over itself.
Either increase the maximum number of allowed simultaneous connections, in
the server's configuration file, or switch to a different mail client that
implements IMAP properly.
"Sam" <s...@email-scan.com> wrote in message
news:cone.1141862165...@commodore.email-scan.com...
Gerard Bok
wrote:
>Gerard Bok writes:
>> Situation: 10 mailboxes on a Courier IMAP server.
>> Multiple users on multiple locations, all sharing all mailboxes.
>> Synchronize accounts produces loads of 0x800CCC0F errors.
>> More specific: at quiet times, I am able to sync each mailbox
>> individually without any error.
>> Once I try to sync all accounts at once, loads of errors are
>> reported, on random mailboxes and even when I just finished an
>> individual sync.
>By mailboxes I presume you are referring to IMAP folders. I doubt that you
>are logging in to ten different IMAP accounts.
Well, good question. But I guess we are :-)
There are 10 different email addresses on the server, each with
it's own mailbox and set of log in codes (and it's own aliases).
>The default Courier-IMAP server configuration permits a maximum of five
>simultaneous connections from the same IP address.
Before we could get this schema going the server's administrator
had to adjust something from 5 upwards. Now it is set to 100
instead. (Due to NAT up to 3 different PCs could be calling the
server from the same IP, so that would make 30 connections.)
>It appears that your mail client opens a separate connection for each IMAP
>folder. The server refuses to accept any more connections, after the first
>five, and then your mail client vomits all over itself.
There definately is going something wrong. Peeking at the
communications shows [TCP Window Full] [TCP Out-of-order]
messages. I never see an indication of a connection being denied
though.
>Either increase the maximum number of allowed simultaneous connections, in
>the server's configuration file, or switch to a different mail client that
>implements IMAP properly.
The first has been done already, I think. And the second is, I am
afraid, not (yet) an option.
Gerard Bok
<jim.pi...@gmail.com> wrote:
>Often, the problem is caused by having an antivirus program settings to scan
>incoming email. Disabling the option does not leave you unprotected and it
>should be disabled. If that does not solve the problem then review the info
>in this KB article for some other ideas:
>
>813514 - OL: Error 0x800ccc0d or 0x800ccc0f When Receiving and Sending
>E-Mail:
>http://support.microsoft.com/?kbid=813514
The problem only occurs with IMAP, not with POP3.
I would expect that to exclude any hardware or communications
related cause.
Further testing shows, that running at a quet time and without
any firewall at all (augh!) let the errors disappear. So I will
start fiddling with the time-out settings.
Thanks Jim. Et merci Jousset!
>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>news:440ec111...@News.Individual.NET...
>> Situation: 10 mailboxes on a Courier IMAP server.
>> Multiple users on multiple locations, all sharing all mailboxes.
>> Synchronize accounts produces loads of 0x800CCC0F errors.
--
Kind regards,
Gerard Bok
Sam
> Sam's reply was as follows (but since he used encryption w/o providing the
> key, you may have missed it):
That's not the reason he might've "missed" it.
The only reason he might've missed it would be if he was using broken
Microsoft E-mail virus propagation shitware, that fails to properly
implement a ten-year old standard for digital signatures.
Robert Aldwinckle
"Jim Pickering" <jim.pi...@gmail.com> wrote in message
news:ua0ZOZyQ...@tk2msftngp13.phx.gbl
...
> Sam's reply was as follows (but since he used encryption w/o providing the
> key, you may have missed it):
Jim,
The others aren't using OE as their newsreaders
so there may not be a problem for them. ; )
A surprise for me is that both Google Groups and
the web interface seem to have no trouble with Sam's replies.
E.g. normally both refuse to show "attachments".
BTW how does WM do with this case? <EG>
FYI
Robert Aldwinckle
---
Jim Pickering
Hope you feel better, now that you've vented about MS. Have a great day.
--
Jim Pickering, MVP, Outlook Express
https://mvp.support.microsoft.com/profile=F9F51EF1-4AE3-4D23-B2D8-1171988A62D6
"Sam" <s...@email-scan.com> wrote in message
news:cone.1141905612....@commodore.email-scan.com...
Jim Pickering
news:%23ItFw65...@TK2MSFTNGP10.phx.gbl...
> (posting to msnews; no access to USENET newsgroups)
>
>
>
> FYI
>
> Robert Aldwinckle
> ---
No idea, Robert, since I've had problems with newsgroup subs in Vista.
Hopefully, the next build will be better but who knows?
--
Jim P.
PA Bear
> A surprise for me is that both Google Groups and
> the web interface seem to have no trouble with Sam's replies.
Hints from headers?...
X-Mailer: http://www.courier-mta.org/cone/
<snip>
Content-Type: multipart/signed;
boundary="=_mimegpg-commodore.email-scan.com-11676-1141905612-0004";
--
~PA Bear
Sam
Further clue: http://www.faqs.org/rfcs/rfc2015.html
Category: Standards Track, October 1996
A ten-year old Internet standard.
For the clue impaired: as a rule, Internet standards are fully backwards
compatible. A compliant mail client is not required to have any encryption
or digital signature-verification capabilities in order to show
digitally-signed messages. For example: Mozilla Thunderbird, out of the
box, displays digitally-signed messages perfectly fine, and no differently
than non-signed mail. This is because it's a MIME formatted message, and
any E-mail client that properly implements MIME will parse digitally-signed
messages as ordinary multipart MIME messages, perhaps with a tiny
attachment. Mozilla Thunderbird even has a tiny bit of additional logic of
suppressing the attachment that holds the digital signature. Even though
Thunderbird needs a PGP plugin to handle MIME-PGP mail, without the plugin,
and because Thunderbird properly implements MIME, unlike Outlook, it
displays the contents of signed messages without making a big fuss.
Where Microsoft screwed up, and has been incapable of fixing, for the last
ten years, is explained by section 7.2.6 of RFC 1521 (RFC 2046 eventually
superceded RFC 1521, but RFC 1521 was in effect at the time):
7.2.6. Other Multipart subtypes
Other multipart subtypes are expected in the future. MIME
implementations must in general treat unrecognized subtypes of
multipart as being equivalent to "multipart/mixed".
Of course, complying with Internet standards is not a top priority for
Microsoft. If Microsoft's virus distribution software was correctly
designed, in accordance with Internet standards, even if it chooses not
implement digitally-signed messages, which use multipart/signed, it should
treat them exactly as "multipart/mixed" messages, which are ordinary
messages, with ordinary attachments.
Interestingly enough, Microsoft's bugware has extensive support for
multipart/related MIME type (RFC 2387), which was introduced in 1998, two
years after multipart/signed.
So, for the last ten years, clueless MS cheerleaders have been constantly
whining because Outlook Express vomits all over itself when it sees a signed
message. They always blame everyone else when the real problem is the buggy
crapware they're using to occasionally read E-mail (when it's not otherwise
busy sending viruses and trojans to all E-mail addresses pulled from the
address book).
PA Bear
"Sam" <s...@email-scan.com> wrote in message
news:cone.1141949262....@commodore.email-scan.com...
Jim Pickering
sending digitally signed messages. Feeling insecure in a world where you
can't trust anyone? Life must be full of problems for you when you have to
uncover all sorts of rocks to find problems where none exist.
And FWIW, I use Outlook Express and do not have a problem reading your
"replies" or "original posts." But that wouldn't fit in with your Microsoft
is evil and Mozilla is great, would it?
--
Jim Pickering, MVP, Outlook Express
https://mvp.support.microsoft.com/profile=F9F51EF1-4AE3-4D23-B2D8-1171988A62D6
"Sam" <s...@email-scan.com> wrote in message
news:cone.1141949262....@commodore.email-scan.com...
Sam
> Absolutely none of your diabtribe explains to anyone why you insist on
> sending digitally signed messages.
Because I can.
> And FWIW, I use Outlook Express and do not have a problem reading your
> "replies" or "original posts." But that wouldn't fit in with your Microsoft
> is evil and Mozilla is great, would it?
Unfortunately, there's more than one version of Outlook Express out there.
Gerard Bok
<jim.pi...@gmail.com> wrote:
>Often, the problem is caused by having an antivirus program settings to scan
>incoming email. Disabling the option does not leave you unprotected and it
>should be disabled. If that does not solve the problem then review the info
>in this KB article for some other ideas:
>
>813514 - OL: Error 0x800ccc0d or 0x800ccc0f When Receiving and Sending
>E-Mail:
>http://support.microsoft.com/?kbid=813514
Update: By no means cured :-)
Fiddling with the timeout settings as Q813514 suggests does not
help. (And that comes to no surprise at all: the default timeout
is already a full minute. While total sync requires less than a
minute, normally.)
Next thing I tried: on a PC with only SP2's firewall: create an
exception for port 143. That should solve firewall related
problems, I suppose. But it doesn't.
Yet, totally switching off the firewall on that same PC allows me
to sync without any errors :-)
Now, for testing purposes, running a PC without a personal
firewall is OK. But for an office PC, that wouldn't be a real
option, would it ?
On a hunch, that XP Home's 5 connection limit might be
responsible, I installed XP Prof, on the very same machine.
Oddly enough, things don't get better. Far from it!
Even without any firewall or virusscanner I cannot even
synchronize an individual account without frequent errors.
So, having more connections available definitely make things
worse. Does that make any sense ?
May I tap once more on your knowledgable opinions?
>"Gerard Bok" <bok...@zonnet.nl> wrote
>> Situation: 10 mailboxes on a Courier IMAP server.
>> Multiple users on multiple locations, all sharing all mailboxes.
>> Synchronize accounts produces loads of 0x800CCC0F errors.
>> Once I try to sync all accounts at once, loads of errors are
>> reported, on random mailboxes and even when I just finished an
>> individual sync.
--
Kind regards,
Gerard Bok
Ron Sommer
Tools, Accounts, Mail tab, Properties, Advanced tab, ...
Are you sending and receiving on port 143?
--
Ron Sommer
Gerard Bok
<rso...@nospam.ktis.net> wrote:
>>>"Gerard Bok" <bok...@zonnet.nl> wrote
>>
>>>> Situation: 10 mailboxes on a Courier IMAP server.
>>>> Multiple users on multiple locations, all sharing all mailboxes.
>>
>>>> Synchronize accounts produces loads of 0x800CCC0F errors.
>>>> Once I try to sync all accounts at once, loads of errors are
>>>> reported, on random mailboxes and even when I just finished an
>>>> individual sync.
>Tools, Accounts, Mail tab, Properties, Advanced tab, ...
>Are you sending and receiving on port 143?
Yes.
Gerard Bok
<rso...@nospam.ktis.net> wrote:
>>>"Gerard Bok" <bok...@zonnet.nl> wrote
>>
>>>> Situation: 10 mailboxes on a Courier IMAP server.
>>>> Multiple users on multiple locations, all sharing all mailboxes.
>>
>>>> Synchronize accounts produces loads of 0x800CCC0F errors.
>>>> Once I try to sync all accounts at once, loads of errors are
>>>> reported, on random mailboxes and even when I just finished an
>>>> individual sync.
>Tools, Accounts, Mail tab, Properties, Advanced tab, ...
>Are you sending and receiving on port 143?
Yes.
Or, to be more precise: IMAP communicates on port 143.
(SMTP still uses port 25. But there is never any mail in the
outbox. So that cannot cause any of the errors.)
use...@isbd.co.uk
>
> Now, for testing purposes, running a PC without a personal
> firewall is OK. But for an office PC, that wouldn't be a real
> option, would it ?
>
sort of half decent router then that should provide the firewall
anyway.
--
Chris Green
Jim Pickering
LinkSys router? Have you searched for any updated firmware for it from the
manufacturer?
It sounds like it could be a configuration problem in your OE accounts. Let
OE make a connection, then click Tools|Accounts|Mail. In the connection
column, they should all show "Any Available". If not, double-click that
account, click Connection, then remove the check mark for "Always connect
using...". Then do the same for any news accounts you might have. OE will
then pass all control for Internet connections to Internet Explorer, and so
will use whatever configuration you have under Control Panel| Internet
Options| Connection. See if that helps at all.
--
Jim Pickering, MVP, Outlook Express
https://mvp.support.microsoft.com/profile=F9F51EF1-4AE3-4D23-B2D8-1171988A62D6
"Gerard Bok" <bok...@zonnet.nl> wrote in message
news:44118973...@News.Individual.NET...
Gerard Bok
<jim.pi...@gmail.com> wrote:
>Is there by any chance a router involved in the total equation? Is it a
>LinkSys router? Have you searched for any updated firmware for it from the
>manufacturer?
The router involved is a PC, running Smoothwall (with patches 1
to 8).
>It sounds like it could be a configuration problem in your OE accounts.
Quite unlikely, as the problems occur at all locations.
Accounts are created on different PC's, different locations,
different ISP's , etc.
The only thing they have in common is the administrator.
(And he is beyond any suspicion. Anyone know the smiley for an
aureole ?)
> Let
>OE make a connection, then click Tools|Accounts|Mail. In the connection
>column, they should all show "Any Available".
Check :-)
>See if that helps at all.
No! Sorry :-) But I am very grateful for all suggestions though.
Also very weird is the fact, that there is a link with the
server's load. But although I can set the timeout value at 3
minutes, the whole circus crashes in a session that lasts 20
seconds in total.
At 5 PM, I can hardly get any mail at all. At 8 PM I can.
Without firewall, all accounts at once. With the firewall
enabled, just one by one.
>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>news:44118973...@News.Individual.NET...
>> On Wed, 8 Mar 2006 11:59:10 -0800, "Jim Pickering"
>> <jim.pi...@gmail.com> wrote:
>>>813514 - OL: Error 0x800ccc0d or 0x800ccc0f When Receiving and Sending
>>>E-Mail:
>>>http://support.microsoft.com/?kbid=813514
>>>> Situation: 10 mailboxes on a Courier IMAP server.
Jim Pickering
similar problems. Does the mail provider have any support forums or
technical assistance for you to go to? It truly sounds like a firewall
problem but you say you have the appropriate port (143) opened but I wonder
if something else is interfering such as an antivirus or antispam program?
Beyond that, I'm stumped unless you have some "malware" installed that is
interfering.
--
Jim Pickering, MVP, Outlook Express
https://mvp.support.microsoft.com/profile=F9F51EF1-4AE3-4D23-B2D8-1171988A62D6
"Gerard Bok" <bok...@zonnet.nl> wrote in message
news:4411d2aa...@News.Individual.NET...
Gerard Bok
<jim.pi...@gmail.com> wrote:
>Gerard: It's hard for me to believe that you are the only user having
>similar problems. Does the mail provider have any support forums or
>technical assistance for you to go to?
They reply that 0x800CCC0Fdoes not indicate any server site
problem but is probably caused by a single nasty email message.
>It truly sounds like a firewall
>problem but you say you have the appropriate port (143) opened but I wonder
>if something else is interfering such as an antivirus or antispam program?
No. I can see (Ethereal) that the communication gets mixed up.
And in the seconds range, not in the minute or more that OE uses
for a time out.
>Beyond that, I'm stumped unless you have some "malware" installed that is
>interfering.
Well, this is what I did (today): A fresh install of Windows XP
Prof, SP2. No virusscanner. No firewall, other than SP2.
(No updates either. ) Actually only installed it to see if Home's
connections limit plays a part.
I cannot imagine a cleaner system than this one :-)
One other thing I tried: I installed Thunderbird.
But that doesn't run at all. After a few mailboxes it says "you
need to extend the maximum number of connections" But it keeps
telling me that, even if I increase the value from 5 to 30.
Next move, under consideration is: transfer the entire mail
system from Courier to Squirrel. And see how that works out.
Jim Pickering
knowing how you did it.
Best regards.
--
Jim Pickering
"Gerard Bok" <bok...@zonnet.nl> wrote in message
news:4412115...@News.Individual.NET...
Robert Aldwinckle
> On Fri, 10 Mar 2006 14:28:30 -0800, "Jim Pickering"
> <jim.pi...@gmail.com> wrote:
>
>>Gerard: It's hard for me to believe that you are the only user having
>>similar problems. Does the mail provider have any support forums or
>>technical assistance for you to go to?
>
> They reply that 0x800CCC0Fdoes not indicate any server site
> problem but is probably caused by a single nasty email message.
Have you tried using the IMAP.log (OE troubleshooting log for IMAP)
to try to see what all that code might be representing?
>
>>It truly sounds like a firewall
>>problem but you say you have the appropriate port (143) opened but I wonder
>>if something else is interfering such as an antivirus or antispam program?
>
> No. I can see (Ethereal) that the communication gets mixed up.
> And in the seconds range, not in the minute or more that OE uses
> for a time out.
Can you tell from that the source of the "mix-up"?
E.g., any sign of Sam's scenario occurring?
Good luck
Robert
---
Bill Kearney
> But that doesn't run at all. After a few mailboxes it says "you
> need to extend the maximum number of connections" But it keeps
> telling me that, even if I increase the value from 5 to 30.
That thunderbird also fails says there's something wrong on the server-side,
or the router link into it. OE is generally pretty stable, as is Tbird.
Outlook's IMAP implementation just sucks.
Yizhar Hurwitz
>
> The aim is to use Outlook 2003.
> Couldn't get that operational at all, so for the time being, we
> settled for Outlook Express instead.
>
HI.
If disabling Windows XP Firewall solves the problem, then you can/should:
Disable it if not needed - anyway you should normally use a network perimeter firewall.
Or re investigate the exceptions you try to use.
You can try exceptions by application exe file name (Add Program), instead of tcp port.
Please note that exception by port number at the client side is irrelevant because such exceptions are intended for server side
applications.
Another suggestion - you have mentioned that you are running "Courier imap" server which uses Maildir format.
Therefor, another possible troubleshooting step might be:
install a different imap server that supports Maildir format, and test how it performs.
Or better and safer test - install a totally different server on different hardware beside the real one, and put some test
mailboxes on it, then configure one of the client to work against that server and check.
Another thing to test is your firewall - it might be related to the problems that packets come out of order and this might
indicate a lower layer (OSI layer 2-4) related problem .
One way to rule this out is simply to use a different firewall from the many options out there.
Again - you can implement a parallel firewall beside the "real" one, and configure a specific PC to use the testing firewall as
default gateway, without affecting the rest of the network.
What do you think?
Yizhar Hurwitz
http://yizhar.mvps.org
Gerard Bok
<yiz...@mail.dot.com> wrote:
>Gerard Bok wrote:
>>
>> The aim is to use Outlook 2003.
>> Couldn't get that operational at all, so for the time being, we
>> settled for Outlook Express instead.
>>
>> Synchronize accounts produces loads of 0x800CCC0F errors.
>If disabling Windows XP Firewall solves the problem, then you can/should:
>
>Disable it if not needed - anyway you should normally use a network perimeter firewall.
It's not just Windows' firewall, but any personal firewall.
There is a perimeter firewall (Smoothwall) but I wouldn't like
the PC's to run without protection of their own.
>Or re investigate the exceptions you try to use.
>You can try exceptions by application exe file name (Add Program), instead of tcp port.
>Please note that exception by port number at the client side is irrelevant because such exceptions are intended for server side
>applications.
You're quite right there. I feel ashamed, as I should have
realized that before I opened port 143.
>Another suggestion - you have mentioned that you are running "Courier imap" server which uses Maildir format.
>Therefor, another possible troubleshooting step might be:
>install a different imap server that supports Maildir format, and test how it performs.
I cannot run another mailserver on that domain as it is hosted by
a third party.
I am currently testing on another domain though, but it appears
to run the same type of Courier mailserver. (Squirrel is just the
webmail frontend, didn't know that :-)
>Another thing to test is your firewall - it might be related to the problems that packets come out of order and this might
>indicate a lower layer (OSI layer 2-4) related problem .
I just connected a PC in front of the firewall. With it's own
firewall also switched off, it still produces errors. Even on a
quiet sunday afternoon. (Server load seems to aggregate the
problems.)
>One way to rule this out is simply to use a different firewall from the many options out there.
>Again - you can implement a parallel firewall beside the "real" one, and configure a specific PC to use the testing firewall as
>default gateway, without affecting the rest of the network.
>
>What do you think?
That you provided me with a lot of new tasks :-)
But I appreciate your input. Anything to get the client happy :-)
Gerard Bok
<rob...@techemail.com> wrote:
>(posted from msnews; no access to USENET)
>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>news:4412115...@News.Individual.NET...
>> On Fri, 10 Mar 2006 14:28:30 -0800, "Jim Pickering"
>> <jim.pi...@gmail.com> wrote:
>>
>>>Gerard: It's hard for me to believe that you are the only user having
>>>similar problems. Does the mail provider have any support forums or
>>>technical assistance for you to go to?
>>
>> They reply that 0x800CCC0Fdoes not indicate any server site
>> problem but is probably caused by a single nasty email message.
>
>Have you tried using the IMAP.log (OE troubleshooting log for IMAP)
>to try to see what all that code might be representing?
Actually, I hadn't.
But now that I have checked both the email and the IMAP box: no
log file gets created at al !
There is also noting relevant in the event viewer.
>>>It truly sounds like a firewall
>>>problem but you say you have the appropriate port (143) opened but I wonder
>>>if something else is interfering such as an antivirus or antispam program?
Well, my stupidity was certainly interfering :-)
As Yitzak pointed out, it's no use opening port 143.
But even without any firewall the problems persist.
>> No. I can see (Ethereal) that the communication gets mixed up.
>> And in the seconds range, not in the minute or more that OE uses
>> for a time out.
>
>Can you tell from that the source of the "mix-up"?
Still puzzling on that one :-)
>
>E.g., any sign of Sam's scenario occurring?
Well, problems surely start with more than 5 mailboxes.
I also got word from the ISP involved: they did not raise the
number of simultaneous connections on that mail account.
But that shouldn't matter as XP Home is not able to make more
than 5 connections anyway, I think.
And problems also occur when there's only one single user active.
Yizhar Hurwitz
>
> I also got word from the ISP involved: they did not raise the
> number of simultaneous connections on that mail account.
> But that shouldn't matter as XP Home is not able to make more
> than 5 connections anyway, I think.
> And problems also occur when there's only one single user active.
>
You are wrong.
The connection limit in Windows XP (both Home and Pro versions) with SP2,
refers to embryonic or half-open connections when the XP clients sends the first packet of the TCP handshake,
and is waiting for the first response from the server (or whatever device in the other side).
If the TCPIP stack on the XP computer reaches that limit, you will find a yellow Event ID 4226 that tells you.
So, if you don't see Event 4226, then this is not the issue.
You can read more info about this limit and workaround here:
SpeedGuide.net :: Windows XP SP2:
http://www.speedguide.net/read_articles.php?id=1497
Yizhar Hurwitz
http://yizhar.mvps.org
Gerard Bok
<yiz...@mail.dot.com> wrote:
>Gerard Bok wrote:
>>
>> I also got word from the ISP involved: they did not raise the
>> number of simultaneous connections on that mail account.
>> But that shouldn't matter as XP Home is not able to make more
>> than 5 connections anyway, I think.
>> And problems also occur when there's only one single user active.
>You are wrong.
Could be :-)
>The connection limit in Windows XP (both Home and Pro versions) with SP2,
>refers to embryonic or half-open connections when the XP clients sends the first packet of the TCP handshake,
>and is waiting for the first response from the server (or whatever device in the other side).
>
>If the TCPIP stack on the XP computer reaches that limit, you will find a yellow Event ID 4226 that tells you.
>So, if you don't see Event 4226, then this is not the issue.
>
>You can read more info about this limit and workaround here:
>SpeedGuide.net :: Windows XP SP2:
>http://www.speedguide.net/read_articles.php?id=1497
I think this story is rather different from Q314882 which talks
about "Any file, print, named pipe, or mail slot session...."
Aren't we talking about mail slot sessions here ?
Sam
Correct.
As I explained before, the IMAP server's default settings impose a limit of
5 connections from the same IP address. You have confirmed with your host
that the server's default settings have not been adjusted. So that's your
answer.
Poorly-written Microsoft bugware sometimes opens a separate connection for
each folder in the mailbox. With large mailboxes, Outhouse was observed
attempting to create hundreds of IMAP connections at the same time.
Creating a new IMAP connection, and logging in, is a relatively expensive
process. Passwords must be validated, and memory must be allocated for each
new IMAP login, to hold the related server process information and mailbox
data. Many systems place account and password information on a separate,
hardened server, which must be accessed for each login attempt. Server RAM
is a limited resource. When encryption is available, SSL
encryption/decryption is CPU intensive, and ain't cheap either.
So, to proactively prevent server-wide resource shortages due to a
denial-of-service attack from poorly-designed software, the number of
default incoming connections from the same IP address is limited to five.
So, you'll just have to wait until Microsoft hires some programmers who have
enough knowledge and experience to implement IMAP correctly. Of course,
robust implementation and support of free, non-proprietary, commodity
Internet standards is not likely to rank a priority for Microsoft.
To pass some time, while waiting for the pigs to fly overhead, you might
want to try to get an answer from their fan club why their virus
distribution system can't live without a separate IMAP connection for each
mailbox. What possible "innovation" demands such pig-headed behavior. I
can do everything Outhouse does, IMAP-wise, and I manage to do it all with a
single IMAP connection, no matter how many folders there are in the mailbox.
This is not rocket science. This is simply sloppy programming on
Microsoft's part.
Gerard Bok
wrote:
True. But in this case we cannot blame MS.
These 10 mailboxes are indeed 10 separate user accounts, each
with it's own set of access codes.
(I just answered the ISP's counter question "what do you mean by
subaddressing?" :-)
Any length to solve / avoid the problem :-)
Yizhar Hurwitz
> I think this story is rather different from Q314882 which talks
> about "Any file, print, named pipe, or mail slot session...."
>
> Aren't we talking about mail slot sessions here ?
Nop.
Q314882 talks about inbound connections, but this discussion is about outbound sessions.
Q314882 talks about MS protocols (SMB), this is not the case here.
Yizhar Hurwitz
http://yizhar.mvps.org
Gerard Bok
<rob...@techemail.com> wrote:
OK. This is the story. I did my homework on a grabbed session :-)
The session lasted some 20 seconds, I grabbed 3480 packets
(100%).
I have 14 TCP connections that result in IMAP communication.
(And 4 connections for which I can find no function. They are
initiated by the host and go SYN SYN/ACK ACK FIN/ACK ACK FIN/ACK
ACK, without any data.)
Basically, I see TCP sessions being build, IMAP log-in, some data
transfer, Idle request, Server going idle, TCP session closed
down. Seems decent to me.
In about half the TCP sessions, the server transmits one or more
[TCP Window Full] [TCP Out-of-Order] messages after the session
was closed.
Am I correct in assuming that it is these packets that send OE
haywire ?
Odd situation:
Port 1038 opens and closes an account; packets 161 - 179
Port 1039 opens the same account again, packets 180 - 397.
The first session has no errors. In the latter some mail gets
downloaded,
host requests Idle, Server enters idle, session is closed.
After that, the server sends a [TCP Window Full] [TCP
Out-of-Order] message.
One other account:
Open on port 1046 in packet 3408
That same account gets opened on port 1047 in packet 3426
Thus, while the account still has an open session on port 1046
Port 1047 closes the account in packet 3452 without an error
Port 1046 closes in packet 3460 with an error
Is this 'embraced opening' proper behaviour ?
Sam
> On Sat, 11 Mar 2006 05:50:14 -0500, "Robert Aldwinckle"
> <rob...@techemail.com> wrote:
>
>>(posted from msnews; no access to USENET)
>>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>>news:4412115...@News.Individual.NET...
>>> On Fri, 10 Mar 2006 14:28:30 -0800, "Jim Pickering"
>>> <jim.pi...@gmail.com> wrote:
>
>>>>It truly sounds like a firewall
>>>>problem but you say you have the appropriate port (143) opened but I wonder
>>>>if something else is interfering such as an antivirus or antispam program?
>>>
>>> No. I can see (Ethereal) that the communication gets mixed up.
>>> And in the seconds range, not in the minute or more that OE uses
>>> for a time out.
>
>>Can you tell from that the source of the "mix-up"?
>>
>>E.g., any sign of Sam's scenario occurring?
>
> OK. This is the story. I did my homework on a grabbed session :-)
> The session lasted some 20 seconds, I grabbed 3480 packets
> (100%).
>
> I have 14 TCP connections that result in IMAP communication.
> (And 4 connections for which I can find no function. They are
> initiated by the host and go SYN SYN/ACK ACK FIN/ACK ACK FIN/ACK
> ACK, without any data.)
If you compare the timestamps, you should find that these 4 connections were
initiated when you already have 5 open connections, your maximum limit. As
soon as the server sees an attempted 6th connection, it gets closed without
further notice. The "SYN SYN/ACK ACK FIN/ACK ACK FIN/ACK ACK" sequence is
exactly what you would expect to see.
> Basically, I see TCP sessions being build, IMAP log-in, some data
> transfer, Idle request, Server going idle, TCP session closed
> down. Seems decent to me.
> In about half the TCP sessions, the server transmits one or more
> [TCP Window Full] [TCP Out-of-Order] messages after the session
> was closed.
> Am I correct in assuming that it is these packets that send OE
> haywire ?
No, these packets are, in most likelyhood, get discarded with no further
processing. OE has a conniption fit when the server takes a new connection
(SYN SYN/ACK ACK) and then immediately closes it (the FIN packets) with no
further response. After a new connection is established OE expects the
server to send the initial IMAP greeting. The server closes the connection
instead, and OE has a cow -- an unexpected error condition that OE's sloppy
programming did not anticipate.
Bruce Hagen
Do not send replies to posts in attachment form, (I won't open them), and
include all original messages in your replies.
--
Bruce Hagen
MS MVP - Outlook Express
~IB-CA~
"Sam" <s...@email-scan.com> wrote in message
news:cone.114221632...@commodore.email-scan.com...
Sam
> Sam,
>
> Do not send replies to posts in attachment form, (I won't open them), and
Bruce,
Do not post to Usenet newsgroups on messaging related subjects unless you
have even the slightest clue regarding Internet messaging standards. I
don't post any messages in "attachment form", and it's not my fault you are
using Microsoft shitware that fails to comply with RFC 2015 - a ten year old
Internet standard.
> include all original messages in your replies.
I do. You need to stop drinking Microsoft kool-aid.
> --
> Bruce Hagen
> MS MVP - Outlook Express
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!
If that doesn't say it all -- an "Outlook Express MVP" who is absolutely
clueless on the subject of Internet messaging standards.
Maybe, dear "Outlook Express MVP", maybe when you've actually written even
one line of code that's makes even the tiniest, slightest, contribution
towards handling E-mail or Usenet-related tasks, then perhaps you can
lecture me on the subject of messaging formats.
But not until then.
Bruce Hagen
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA
HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!
If that doesn't say it all -- an "Outlook Express MVP" who is absolutely
clueless on the subject of Internet messaging standards.
Maybe, dear "Outlook Express MVP", maybe when you've actually written even
one line of code that's makes even the tiniest, slightest, contribution
towards handling E-mail or Usenet-related tasks, then perhaps you can
lecture me on the subject of messaging formats.
But not until then.
***********************************
Typical Massachusetts Liberal Jerk. Welcome to my Killfile. I will not be
bothered by your posts any longer as I will no see them again.
Here's my one line of code:
X-Trace:
sv3-DjoFnHsrg7AbLCx60MeDi3wskvtuVTv2VBlr2hPWtY1n/DtYwQduSNh/bZ4ZZMxyHajxLIyTqvlDUSn!acDROiiSXPPC8ChLWBcKC0JouyZqn4gGPHNC0qgmKiSx3rMMo2aLgNadAJUGwU5rm3KTTzxfnJ9m!RWDtVGZHOIN5zBW1p87PUPVFDmSypsi8u+f747TIzGVJ
X-Complaints-To: ab...@speakeasy.net
X-DMCA-Complaints-To: ab...@speakeasy.net
X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers
X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint
properly.
Good Newsgroup posts:
http://dts-l.org/goodpost.htm
Goodbye.
--
Bruce Hagen
MS MVP - Outlook Express
~IB-CA~
"Sam" <s...@email-scan.com> wrote in message
news:cone.1142220250...@commodore.email-scan.com...
Robert Aldwinckle
>>Have you tried using the IMAP.log (OE troubleshooting log for IMAP)
>>to try to see what all that code might be representing?
>
> Actually, I hadn't.
> But now that I have checked both the email and the IMAP box: no
> log file gets created at al !
That's a surprise. Unfortunately I don't have an IMAP account
so I'm only guessing that IMAP.log is the name of the troubleshooting
log file created when IMAP is checked in the Options, Maintenance tab.
I do know that SMTP.log is created (or added to) when Mail is checked
there. AFAIK all logs are created in the OE Message Store directory.
Is that where you are looking for them? Can you see other logs in that
same directory? E.g. if you activate troubleshooting logging for News
and do a POST or a Refresh in a newsgroup you should see a log
with the name of that newsgroup.
HTH
Robert
---
Robert Aldwinckle
> On Sat, 11 Mar 2006 05:50:14 -0500, "Robert Aldwinckle"
> <rob...@techemail.com> wrote:
>
>>(posted from msnews; no access to USENET)
>>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>>news:4412115...@News.Individual.NET...
>>> No. I can see (Ethereal) that the communication gets mixed up.
>>> And in the seconds range, not in the minute or more that OE uses
>>> for a time out.
>
>>Can you tell from that the source of the "mix-up"?
>>
>>E.g., any sign of Sam's scenario occurring?
>
> OK. This is the story. I did my homework on a grabbed session :-)
> The session lasted some 20 seconds, I grabbed 3480 packets
> (100%).
>
> I have 14 TCP connections that result in IMAP communication.
> (And 4 connections for which I can find no function. They are
> initiated by the host and go SYN SYN/ACK ACK FIN/ACK ACK FIN/ACK
> ACK, without any data.)
>
> Basically, I see TCP sessions being build, IMAP log-in, some data
> transfer, Idle request, Server going idle, TCP session closed
> down. Seems decent to me.
> In about half the TCP sessions, the server transmits one or more
> [TCP Window Full] [TCP Out-of-Order] messages after the session
> was closed.
> Am I correct in assuming that it is these packets that send OE
> haywire ?
I don't know (though I find Sam's interpretation believable.)
This is why I was hoping that the IMAP.log would have enough detail
to show you more clearly what OE is seeing and how it is interpreting that.
Some of the stuff that you are describing looks to be at too low a level
to be something that OE would be explicitly requesting. E.g. I suspect
that a lot of it would be generated by higher level API functions.
I would expect the IMAP.log to be displaying information about those
higher level functions and mapping lower level symptoms into negative
return codes such as the one you see being reported.
>
> Odd situation:
> Port 1038 opens and closes an account; packets 161 - 179
> Port 1039 opens the same account again, packets 180 - 397.
> The first session has no errors. In the latter some mail gets
> downloaded,
> host requests Idle, Server enters idle, session is closed.
> After that, the server sends a [TCP Window Full] [TCP
> Out-of-Order] message.
>
> One other account:
> Open on port 1046 in packet 3408
> That same account gets opened on port 1047 in packet 3426
> Thus, while the account still has an open session on port 1046
> Port 1047 closes the account in packet 3452 without an error
> Port 1046 closes in packet 3460 with an error
>
> Is this 'embraced opening' proper behaviour ?
I agree with Sam. I think you are seeing a deficiency in OE for this case.
I've seen the same thing happen with news accounts which are controlled
by login. OE creates far more connections than it needs to per server
and can't disconnect them reliably enough so eventually the server thinks
that the user is logged on more times simultaneously than it can allow.
Typically then it just ignores more connection requests for that user
until the first ones timeout.
Looks to me that you may be further ahead figuring out why your
Outlook 2003 option isn't working. I expect that it would have
better awareness of the problems with the IMAP protocol and more control
with it than OE. IMAP in OE (at least as seen in this newsgroup)
is very much the exceptional case. Most OE Mail accounts are POP3 or HTTP.
Good luck
Robert
---
Gerard Bok
wrote:
I would have loved that to be true ('problem solved').
But unfortunately...
I made a table with 4 'slots' and divided all connections accross
them in such a way that there is no overlap. (I will post or mail
the table if you care.)
There is 1 connection that lasts from packet 384 till 3031;
that's pretty much the whole session.
All others fall nicely in 3 other slots; sometimes head to tail,
sometimes with a gap.
> As
>soon as the server sees an attempted 6th connection, it gets closed without
>further notice. The "SYN SYN/ACK ACK FIN/ACK ACK FIN/ACK ACK" sequence is
>exactly what you would expect to see.
Great. But these shorties are all at the end of my grab.
>> Basically, I see TCP sessions being build, IMAP log-in, some data
>> transfer, Idle request, Server going idle, TCP session closed
>> down. Seems decent to me.
>> In about half the TCP sessions, the server transmits one or more
>> [TCP Window Full] [TCP Out-of-Order] messages after the session
>> was closed.
>> Am I correct in assuming that it is these packets that send OE
>> haywire ?
>
>No, these packets are, in most likelyhood, get discarded with no further
>processing. OE has a conniption fit when the server takes a new connection
>(SYN SYN/ACK ACK) and then immediately closes it (the FIN packets) with no
>further response. After a new connection is established OE expects the
>server to send the initial IMAP greeting. The server closes the connection
>instead, and OE has a cow -- an unexpected error condition that OE's sloppy
>programming did not anticipate.
Well, I'm not an expert in the field, but it looks to me as if it
is the server that sends [TCP Window Full] [TCP Out-of-Order] on
a port on which the connection was terminated.
And I think the server shouldn't do that.
FIN/ACK RST/ACK and then, well, that probably not proper language
in this group :-)
Gerard Bok
<rob...@techemail.com> wrote:
>"Gerard Bok" <bok...@zonnet.nl> wrote in message
>news:44143212...@News.Individual.NET...
>> On Sat, 11 Mar 2006 05:50:14 -0500, "Robert Aldwinckle"
>> <rob...@techemail.com> wrote:
>
>>>Have you tried using the IMAP.log (OE troubleshooting log for IMAP)
>>>to try to see what all that code might be representing?
>>
>> Actually, I hadn't.
>> But now that I have checked both the email and the IMAP box: no
>> log file gets created at al !
>
>That's a surprise. Unfortunately I don't have an IMAP account
>so I'm only guessing that IMAP.log is the name of the troubleshooting
>log file created when IMAP is checked in the Options, Maintenance tab.
OK. IMAP log gets created; loads of them. They appear to be
rather effectively hidden though :-)
This session: attempt to get folders for an empty mailbox.
The server is up and and running and allowing sync to (some)
other mailboxes.
Microsoft Internet Messaging API 6.00.2900.2670
(xpsp_sp2_gdr.050504-1643)
IMAP Log started at 03/13/2006 12:43:00
IMAP: 12:43:00 [db] Bezig met het maken van een verbinding naar
mail.domain.nl op poort 143.
(attempting to connect to mail...)
IMAP: 12:43:00 [db] OnNotify: asOld = 0, asNew = 4, ae = 0
IMAP: 12:43:00 [db] OnNotify: asOld = 4, asNew = 5, ae = 2
IMAP: 12:43:00 [db] OnNotify: asOld = 5, asNew = 5, ae = 4
IMAP: 12:43:00 [db] Verbinding met mail.domain.nl verbroken.
(connection with mail.domain.nl closed)
IMAP: 12:43:00 [db] OnNotify: asOld = 5, asNew = 0, ae = 5
IMAP: 12:43:00 [db] ERROR: "De server heeft de verbinding
onverwachts afgebroken. Mogelijke oorzaken
zijn problemen met de server, netwerkproblemen of te lange
inactiviteit.", hr=2148322319
(Error: The server unexpectedly closed the connection.
Possible causes are: problems with the server,
network problems or exceeded inactivity, hr=)
Translation from dutch added. domain.nl is a placeholder.
Sam
> Typical Massachusetts Liberal Jerk.
I don't live in Massachusetts, spanky. Neither am I a liberal. I'm
actually somewhat curious what led you to such a brilliant conclusion. I
guess you have to be a Microsoft MVP in order to posess such amazing
detective skills.
> Welcome to my Killfile. I will not be
> bothered by your posts any longer as I will no see them again.
The pile of sand is over there. Feel free to stick your head into it,
anytime you get the urges.
> Here's my one line of code:
>
> X-Trace:
> sv3-DjoFnHsrg7AbLCx60MeDi3wskvtuVTv2VBlr2hPWtY1n/DtYwQduSNh/bZ4ZZMxyHajxLIyTqvlDUSn!acDROiiSXPPC8ChLWBcKC0JouyZqn4gGPHNC0qgmKiSx3rMMo2aLgNadAJUGwU5rm3KTTzxfnJ9m!RWDtVGZHOIN5zBW1p87PUPVFDmSypsi8u+f747TIzGVJ
> X-Complaints-To: ab...@speakeasy.net
> X-DMCA-Complaints-To: ab...@speakeasy.net
> X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers
> X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint
> properly.
That's not code. But, that kind of thinking is not unexpected, after
drinking Microsoft kool-aid. Confusion between code and data is quite
understandable, and the resulting security implications are obvious. Case
in point: Outlook Express.
> Good Newsgroup posts:
> http://dts-l.org/goodpost.htm
>
> Goodbye.
> --
> Bruce Hagen
> MS MVP - Outlook Express
> ~IB-CA~
Good things to know for any "MVP - Outlook Express".
http://www.faqs.org/rfcs/rfc2015.html
http://www.faqs.org/rfcs/rfc2046.html
Sam
> On Sun, 12 Mar 2006 20:18:41 -0600, Sam <s...@email-scan.com>
> wrote:
>
>>If you compare the timestamps, you should find that these 4 connections were
>>initiated when you already have 5 open connections, your maximum limit.
>
> I would have loved that to be true ('problem solved').
> But unfortunately...
> I made a table with 4 'slots' and divided all connections accross
> them in such a way that there is no overlap. (I will post or mail
> the table if you care.)
It's not always that easy.
In most cases, IMAP clients take the easy way out and just close the
connection. That's the quick and dirty way. Very few invest the time in
doing this correctly: sending the logout command and wait a little bit for
the server to reply; if the connection is encrypted, engage in an orderly
SSL session shutdown; and finally close the network connection. On the
other end the server usually has additional tasks to do, which typically
involves logging the terminated connection, cleaning up any temporary files,
and other tasks. This doesn't take long, but it does take a bit of time.
And until everything is done, that connection is still on the books, and
still counts towards your per-IP limit. It is entirely possible that you
have already moved on and have tried to establish a new connection, but the
old one is still in the final stages of wrapping up.
If you're doing this very quickly, it's perfectly feasible for you to run
into the per-IP limit even if you think you can't be. Especially if you're
doing all this on a local LAN, with negligible client/server latency.
Try putting in a brief delay after you close an old connection, and before
you open a new one. 100 milliseconds should be more than enough.