cancel
Showing results for 
Search instead for 
Did you mean: 

Mail delivery to recipients taking 5hrs plus

Community Veteran
Posts: 1,996
Thanks: 5
Fixes: 5
Registered: ‎01-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

I've read through alot if this, and been a 'victim' of the issues. Whilst generally quite it literate email is not the case.
Can anyone explain a couple of things....
1.. What is the actual issue here causing the problem.
2.. Who is responsible, for providing resolution, plusnet or Google, and what might be the resolution.

I am more than happy to open a case with Google if that helps although I suspect plusnet may be able to open more doors than me.
Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

@penfold

I already have a support question open on Google's support forums … unlike this space, Google support staff do not appear to respond to posts … or will not respond to this one.  In answer to your questions...

  1. The actual cause has not been asserted with clarity by Google.  Emails sent by the same Plusnet user, using the same Plusnet email servers will be accepted straight away if the Gmail recipient pays Google for their mailbox, or acceptance might be delayed between a few to many hours if the user is a free account.  If the Plusnet user uses their own domain name, with the same Plusnet email servers, there are no issues.
  2. What the resolution might be cannot be stated with certainty until there is certainty over the cause.

There have been a number of references to various email domain DNS TXT records.  These would be a general help in the wider context of sender verification etc.  However there are two considerations...

  1. These are not mandatory and in their absence the accepting email regime should behave in a neutral manner
  2. If Google are in fact breaking the neutrality rules, then why does it allow instant email acceptance to paid for accounts, whilst delaying acceptance to free accounts?

Note the empirical evidence suggests this is not a delay in the Plusnet servers sending the email (or attempting to do so) it is a refusal to accept delivery (by the Gmail servers) where the recipient is a free account.  The Plusnet servers then back off for a while and try again later … and repeats until accepted … or five days later will report a failure to deliver.

It might well be the case that the issue could be mitigated by Plusnet implementing DNS TXT records for every username.plus.com domain name … but as yet there is no evidence or confirmation that this will resolve the issue.  Such a change would not be trivial.

In the meantime, Gmail servers behave in a manner which clear indicates that the absence of such records for userowndomain.co.uk is of no relevance.

In short there is speculation that there are remedies in PlusNet's space to work around Gmail's inconsistent behaviour … but such remedies can be seen to be irrelevant to how things are working fine for the paid for Gmail accounts.  That being the case, I would suspect that there is some other, not yet identified issue here, which might well not be fixed by additional DNS TXT records.

What is required is for Google to "come clean" and explain why they are acting differently depending on which of their customers the emails are addressed to.

@outlook.com really does look like a better alternative!

bijou55
Newbie
Posts: 2
Thanks: 2
Registered: ‎18-08-2018

Re: Plusnet to Gmail addresses - what's the problem?

This is just getting ridiculous ! I sent (or tried to send) an email to my daughter at 16.59 on 28/09/18 and it arrived at 17.17 today (02/10/1Cool. How much longer before this problem is sorted and we oldies can be in reliable contact with our families again ?

Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

That is a good question to ask on the Google forums.

barcud
Dabbler
Posts: 20
Thanks: 2
Registered: ‎05-11-2014

Re: Plusnet to Gmail addresses - what's the problem?

When all else is equal, it is only the type of GMail account that determines whether a given PluNet email is delivered in a timely fashion or is held for an indeterminate period.

GMail's choice: If you pay us we will deliver direct, if you are not paying us - who knows.

How fair or unfair is that analysis?

On the one hand with a paid account Gmail will treat emails 'neutrally' on the other hand ..................

Hmm?Huh

sirjestalot
Grafter
Posts: 25
Thanks: 14
Registered: ‎08-09-2018

Re: Plusnet to Gmail addresses - what's the problem?

If it really is the case that Google are deliberately delaying email to free Gmail accounts, but paid accounts receive their email without delay, one wonders then why some emails I sent to my friend's free Gmail account last week were delivered almost immediately, and yet one or two others were delayed up to 48 hours.  It's the sort of 'behaviour' one would expect to be consistent from a system set up to operate that way.  Perhaps they're using uncertainty as a means to encourage people to switch to a paid account...

Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

From empirical experimentation, the delay has been found to be progressive.  The first email to a target address is delivered promptly, the next after a short delay, the third after a longer delay … reportedly up to 48 hours which is within the standard 5 day retry limit to send a mail item algorithms of most email services.

sirjestalot
Grafter
Posts: 25
Thanks: 14
Registered: ‎08-09-2018

Re: Plusnet to Gmail addresses - what's the problem?

If Townman's hypothesis is correct, does this mean that emails from any one sender to any one (free) Gmail address go through a repeating cycle of "No delay" - "Short delay" - "Long delay"?   In which case, if I send two copies of each email to a free Gmail address, would that ensure that at least one of the copies will either deliver promptly or suffer only short delay?  That's kind of self-defeating because if everyone did the same it would double the email traffic to Gmail...

Or have I misunderstood things?  :-)

Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

In the tests I and others have conducted, a series of emails have been sent to the same address, all fairly close together. One arrives promptly the others with varying intervals of delay. The order received is not necessarily the same as that sent.

I have only performed the test twice (one paid for target the other free), therefore I do not have empirical evidence to know if the increasing delay is for a sender-receipent pair or is just predicated on the sender and to any series of free Gmail receiptents.

I do not know if after a ‘quite’ interval the delay is reset.

I have only one free Gmail account (in addition to my paid for one) which I only use for accessing Google apps. Others might have conducted more in-depth testing.
Superuser
Superuser
Posts: 9,926
Thanks: 1,265
Fixes: 71
Registered: ‎06-04-2007

Re: Plusnet to Gmail addresses - what's the problem?


@sirjestalot wrote:

If Townman's hypothesis is correct, does this mean that emails from any one sender to any one (free) Gmail address go through a repeating cycle of "No delay" - "Short delay" - "Long delay"?


@sirjestalot Unfortunately the delay steps are far more numerous than the simple 3-step one suggested there; I haven't recorded all delays I've experienced in testing but there could be 10, or more steps - and as @Townman mentioned they aren't necessarily progressive.

Regarding "reset"; on a previous occasion I'd checked for delays after not sending for 11 days - and experienced a delay of 3 minutes 4 seconds. I've just checked again after not sending for 27 days - this time the delay was 7 minutes 33 seconds. Such delays are short enough to be regarded as "tolerable" but one email  every two weeks (or so) is hardly practicable. Sad

David
jric
Hooked
Posts: 6
Registered: ‎13-08-2018

Re: Plusnet to Gmail addresses - what's the problem?

Sorry I don't agree with your hypothesis Townman as it doesn't reflect my own experience, which seems more random than that? (Last 2 emails to my wife 48hrs32 mins & 15 mins). Also what would be the rationale, assuming it could be done?

I have now also had a little look around Gmail support and it seems that inward delay is a recurring problem reported by Gmail users, and not just with Plusnet. Nothing much seems to happen at Google in the way of practical solutions or fixes, just suggestions for things beyond the average user like using MX Toolbox.

One advisor stated that it wasn't a good idea to think of email as an instant form of communication. Well that's good advice as it would be quicker to use the post for a lot of my messages!

Also, more worryingly, since starting to make enquiries I now find that about 1 in 4 messages to two separate Australian Gmail addresses have not turned up at all. There is no indication of this at either end, no failed delivery messages – they have just disappeared.

So as far as I can see, the problem is that maybe Gmail and its handling capacity is nowhere near as reliable as people seem to think? However, so many of my contacts are using Gmail these days and I'm losing confidence in an email system that seems to be going backward.  

Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

Anyone is free to agree or disagree with the detail. However your comments support the perspective that this is a Gmail issue impacting a wide user base.

The profile of this issue could give rise to emails not being delivered and not reported as failed until 5 days have passed, which is the typical time before throwing in the towel on retrying to get a delivery accepted. That reminds me that some systems do report delivery issues whilst still attempting to send.

The support comment from Google that email delivery is not instant is a cop-out by a rookie support guy. Yes generally that is right, but where one email is addressed to both a paid for account and a free account and the paid for gets it nearly instantaneously and the free one does not ... then the theory of ‘the delivery time is variable’ does not hold water.

That thought gives rise to more questions which I’ll go seek answers to. There is a lot around this issue which does not make sense.
sirjestalot
Grafter
Posts: 25
Thanks: 14
Registered: ‎08-09-2018

Re: Plusnet to Gmail addresses - what's the problem?

Townman wrote that "some systems do report delivery issues whilst still attempting to send".  That's what our system did at my former place of work. It would make repeated attempts at increasing intervals, but if it did not manage to deliver (i.e. couldn't start an SMTP transaction with the remote server) within 24 hours, the sender would get an alert email to that effect.  This would continue for up to 7 days, after which the message would be de-queued and returned to the sender as undeliverable.  However, once a message had been handed over to the remote server, our system treated it as delivered, but there was of course no guarantee that the message had actually been put in the recipient's mailbox at the other end.  One assumes that delayed Plusnet emails are kept on Plusnet's outgoing queues until accepted for delivery by Gmail, rather than being - rather pointlessly - held at Gmail until delivery?

Superuser
Superuser
Posts: 13,047
Thanks: 4,337
Fixes: 26
Registered: ‎22-08-2007

Re: Plusnet to Gmail addresses - what's the problem?

@sirjestalot,

Yes, this is an acceptance of delivery issue.

-----------------

OK the thought which passed through my head above in respect of the empirical testing was "Why does an email sent to both a paid for and a free gmail accounts have different delivery acceptance behaviours … because Plusnet should hand off the email to BOTH accounts in one SMTP submission?".

However, when looking at the detail of the tests performed, it is not that simple.

@gmail.com ( the free account) uses these MX servers

gmail.com       MX preference = 20, mail exchanger = alt2.gmail-smtp-in.l.google.com
gmail.com       MX preference = 30, mail exchanger = alt3.gmail-smtp-in.l.google.com
gmail.com       MX preference = 10, mail exchanger = alt1.gmail-smtp-in.l.google.com
gmail.com       MX preference = 5, mail exchanger = gmail-smtp-in.l.google.com
gmail.com       MX preference = 40, mail exchanger = alt4.gmail-smtp-in.l.google.com

@business.co.uk (the paid for account) uses these MX servers

business.co.uk   MX preference = 10, mail exchanger = aspmx.l.google.com
business.co.uk   MX preference = 20, mail exchanger = alt1.aspmx.l.google.com
business.co.uk   MX preference = 30, mail exchanger = alt2.aspmx.l.google.com
business.co.uk   MX preference = 40, mail exchanger = aspmx2.googlemail.com
business.co.uk   MX preference = 50, mail exchanger = aspmx3.googlemail.com

In other words the paid for and free gmail accounts are apparently using different MX servers, therefore Plusnet hands these off to Gmail as two distinct activities.  Light bulb thought extinguished!!

 

Does anyone have access to a PAID FOR Gmail account in the @gmail.com address space?  Such would be useful in verifying the presumed difference of what is happening here.

 

I just repeated my test scenarios using

  1. webmail as the client
    1. having its address as @mydomain.co.uk (my name)
    2. and an alias as @myaccount.plus.com (native)
  2. sending 2 test mails "from" each of to the two above to BOTH of
    1. @gmail.com (free account)
    2. @business.co.uk (paid for account

For both tests sending as 1.1 above, 2.1 and 2.2 were recieved in their Gmail inboxes at the same time.

For each test sending as 1.2 above, 2.2 arrived before 2.1 with an increasing delay, first 7 minutes, then 15.

@business.co.uk

Capture Gmail paid.JPG

@gmail.com

Capture Gmail free.JPG

Eddie
Hooked
Posts: 8
Thanks: 4
Registered: ‎01-07-2016

Re: Plusnet to Gmail addresses - what's the problem?

149 posts over several months; the cause of the problem is well known to Plusnet and they have not addressed it with Google or by modifying their format; obviously this is going nowhere.

I'm off.

Goodbye plusnet.

In Tibault's words, 'A plague on both your houses.'