Big Pond Mail trouble
Rachel Polanskis
rachel@juno.virago.org.au
Tue, 17 Jun 1997 21:01:11 +1000 (EST)
Hi linkers,
Here's a few snippets from another list I'm on (oz-isp)
regarding problems with Telstra's Big Pond mail subsystem.
Has anyone else heard more about this?
What was the ABC intending to do with the information?
I would have thought Telstra would have more sense than to relegate
a mission critical service such as mail to a broken OS such as NT.
I wonder if this is an indication just how deeply they are intertwined
with the MS monster?
Apologies to those of you who saw these posts already,
and I hope those who see their own posts in here appreciate
why I appropriated them ;)
rachel
From: Kimberley Heitman <kheitman@it.net.au>
To: aussie-isp@aussie.net
Subject: [Oz-ISP] Telstra email debacle
Date: Tue, 17 Jun 1997 13:26:40 +0800 (WST)
I've been asked to talk in interview with the ABC today about the Telstra
email crash , and the views of users and other isps.
Any thoughts from listmembers as to background info , advice or occurences
that might be relevant ?
Kim
----------------------------------------------------------------------
Kimberley James Heitman
http://www.multiline.com.au/~kheit/
Internet kheitman@it.com.au Fidonet 3:690/254.14
Telephone +618 9458 2790 Facsimile +618 9356 1247
----------------------------------------------------------------------
Date: Tue, 17 Jun 1997 15:40:07 +0930 (CST)
From: Stephen Baxter <steve@senet.com.au>
Subject: Re: [Oz-ISP] Telstra email debacle
Hello,
This is the first I have heard of trouble in the Big Pond.
What happened ?
If you have your mail stored on a one or set of hard drives and there is a
problem with that system then chances are you will lose some email.
While the system is down it just gets placed in a que - no real big deal
but due to the rate at which new mail enters the system and is taken out
of it by users regular backup policies on the mail spool are hard to
justify (ie tape based). Drive mirroring and other methods work better
here such as redundant mail storgae systems.
Multiple drives in non RAID clusters can also greatly improve performance
on any system.
> I've been asked to talk in interview with the ABC today about the Telstra
> email crash , and the views of users and other isps.
>
> Any thoughts from listmembers as to background info , advice or occurences
> that might be relevant ?
>
> Kim
Stephen Baxter SE Network Access
SE Network Access http://www.senet.com.au
Direct Internet Access 222 Grote Street
phone : +61 8 8221 5221 Adelaide 5000
fax : +61 8 8221 5220
<http://www.senet.com.au/~steve/pgp.html for Public Key>
Subject: Re: [Oz-ISP] Telstra email debacle
Date: Tue, 17 Jun 1997 16:13:01 +1000 (EST)
From: "Tom Minchin" <tom@interact.net.au>
> I've been asked to talk in interview with the ABC today about the Telstra
> email crash , and the views of users and other isps.
>
> Any thoughts from listmembers as to background info , advice or occurences
> that might be relevant ?
>
BigPond's email is routed through a single NT PC, which also appears to
handle the POP side of things too.
A person who works for support at BigPond (so this is 'word of mouth') said
there are problems with the software, and it's not able to handle
the number of users (ie it crashed when it reached 100,000 users). Unlike
most largish ISPs, BigPond has no secondary mail exchanges, or indeed,
any backup servers to bring online in case of failure.
Typically mail can only last in a waiting state for 3 to 5 days, so it's
unlikely more than 5% of mail was eventually knocked back as undelivered.
However, most users appeared to treat the repeated warnings from sendmail
type MTAs as failure, and repeated their email attempts over and over
again.
A related Mail & BigPond disaster was when TMNS created an MX record
for "bigpond.com" which pointed to a non-existant address. For 24hours,
any mail to BigPond bounced with an "unknown address" - their support
staff was helpless as they (claimed) had no control over their own
network configuration. TMNS made several botched attempts to fix the
zone information, their first few attempts actually fixed the problem,
but they forgot to update the SOA...
--
Tom Minchin - InterACT Technology Group - Phone: (06) 2578333
tom@InterACT.net.au - Canberra Region Internet Solutions - Fax: (06) 2578322
Tech Support - http://www.InterACT.net.au/ - info@InterACT.net.au
Subject: Re: [Oz-ISP] Telstra email debacle
Date: Tue, 17 Jun 1997 17:03:41 +1000 (EST)
From: "Tom Minchin" <tom@interact.net.au>
To date, BigPond have had several episodes where 'mail.bigpond.com' was
not accepting mail (either down or refusing) for 2 to 3 days continuous.
Also, there was the time when TMNS kindly mx'ed 'bigpond.com' to
'mail1.bigpond.com' (which was non-existant in DNS) and weren't able to
fix the problem for 24 hours - ie they fixed the problem but didn't
change the SOA in the BigPond.com zone etc etc.
BigPond only have a single computer running NT to collect their mail
and handle POP (which is very impressive for 100,000 user accounts).
No secondary MX's, no backup servers...
> Hello,
>
> This is the first I have heard of trouble in the Big Pond.
>
> What happened ?
>
>
> If you have your mail stored on a one or set of hard drives and there is a
> problem with that system then chances are you will lose some email.
>
> While the system is down it just gets placed in a que - no real big deal
> but due to the rate at which new mail enters the system and is taken out
> of it by users regular backup policies on the mail spool are hard to
> justify (ie tape based). Drive mirroring and other methods work better
> here such as redundant mail storgae systems.
> Multiple drives in non RAID clusters can also greatly improve performance
> on any system.
>
Date: Tue, 17 Jun 1997 17:12:14 +0930 (CST)
From: Stephen Baxter <steve@senet.com.au>
Subject: Re: [Oz-ISP] Telstra email debacle
> BigPond only have a single computer running NT to collect their mail
> and handle POP (which is very impressive for 100,000 user accounts).
> No secondary MX's, no backup servers...
To say the least - what sort of RAID controller does it use ?
Is it single/dual/quad processor or is it an Alpha ?
I would really love to know the the IO subsystem as that was where our
mail server was getting stuffed on until we gave it another SCSI channel.
> > Hello,
> >
> > This is the first I have heard of trouble in the Big Pond.
> >
> > What happened ?
> >
> >
> > If you have your mail stored on a one or set of hard drives and there is a
> > problem with that system then chances are you will lose some email.
> >
> > While the system is down it just gets placed in a que - no real big deal
> > but due to the rate at which new mail enters the system and is taken out
> > of it by users regular backup policies on the mail spool are hard to
> > justify (ie tape based). Drive mirroring and other methods work better
> > here such as redundant mail storgae systems.
> > Multiple drives in non RAID clusters can also greatly improve performance
> > on any system.
> >
Stephen Baxter SE Network Access
SE Network Access http://www.senet.com.au
Direct Internet Access 222 Grote Street
phone : +61 8 8221 5221 Adelaide 5000
fax : +61 8 8221 5220
<http://www.senet.com.au/~steve/pgp.html for Public Key>
From: admin@bondi.net (Michael Wines)
Subject: Re: [Oz-ISP] Telstra email debacle
Date: Tue, 17 Jun 1997 08:20:03 GMT
On Tue, 17 Jun 1997 13:26:40 +0800 (WST), you wrote:
:I've been asked to talk in interview with the ABC today about the Telstra
:email crash , and the views of users and other isps.
:
:Any thoughts from listmembers as to background info , advice or occurences
:that might be relevant ?
http://www.bigpond.com/scripts/read_soapbox.idc
Read with a grain of salt, and enjoy.
----
Email "unsubscribe aussie-isp" to majordomo@aussie.net to be removed.
Reminder: posts to list only accepted from your subscribed email address
defghijklmnopqrstuvwxyz: What? No ABC?
--
Rachel Polanskis Kingswood, Greater Western Sydney, Australia
grove@zeta.org.au http://www.zeta.org.au/~grove/grove.html
r.polanskis@nepean.uws.edu.au http://www.nepean.uws.edu.au/ccd/
Witty comment revoked due to funding cuts