FM down...

[vivil]

Server outage since a few minutes, what is the trouble 4 today ? ^^

[Ladrillo]

Down... +10

[bbrener]

Even the home page won't load. Server's not responding .. must be bad

[vivil]

MX 20 in2.smtp.messagingengine.com. 10 in1.smtp.messagingengine.com.


"telnet in1.smtp.messagingengine.com 25" => timeout
"telnet in2.smtp.messagingengine.com 25" => timeout

not good !!!

[disneylogic]

(yeah, I know, bad)

Out. Checked by trying to sign in through two separate networks, and checked pingdom. Does say it's down.

BTW, I saw the brief quirk where some of the Javascript-like actions weren't happening. I thought it was my browser. (Now Google Chrome, at least at home.)

- Jan

[vivil]

Network issues
September 6, 2011 7:01 pm — fastmailstatus

Status (by brong at Tue Sep 6 19:01 UTC)
We are experiencing network issues that mean the site is offline for all users. We will update as soon as we know what’s happening.




edit by vivil: 19:19 UTC. good now.... MX and website works now...

[veloster]

I've been a paying Fastmail customer with multiple accounts for years. I was concerned when Fastmail sold out to Opera. And it seems my concerns may have been justified. Since the change in ownership the reliability of Fastmail has taken a very significant turn for the worse.

Many of us have our business email with Fastmail. It's not acceptable when the free services have much better uptime than a paid service like Fastmail.

Another signifigant outage this year will be the end of all our Fastmail accounts. Email is just too important and this outage is costing real money.

It's also shocking in the other thread here today a Fastmail employee initially denied there was even a problem. Please give your users a little more credit.

I'm not sure what's changed with Fastmail's management, hardware, budget, or whatever, but it needs to be fixed ASAP. This isn't acceptable.

[EdinwolfPA]

Back up for me in PA as of Tuesday, September 6th at 3:30pm EST

David

[brong]

Originally Posted by veloster

I've been a paying Fastmail customer with multiple accounts for years. I was concerned when Fastmail sold out to Opera. And it seems my concerns may have been justified. Since the change in ownership the reliability of Fastmail has taken a very significant turn for the worse.

Many of us have our business email with Fastmail. It's not acceptable when the free services have much better uptime than a paid service like Fastmail.

Another signifigant outage this year will be the end of all our Fastmail accounts. Email is just too important and this outage is costing real money.

It's also shocking in the other thread here today a Fastmail employee initially denied there was even a problem. Please give your users a little more credit.

I'm not sure what's changed with Fastmail's management, hardware, budget, or whatever, but it needs to be fixed ASAP. This isn't acceptable.

It's one particular type of new hardware that has now had a second failure in exactly the same way, and we have further debugging information this time that we can take back to the vendor and ask for more information.

Er - denying there's a problem 12 hours beforehand, when there actually wasn't a problem. I have no issue with that.

Downtime does suck, and we hate it as much as anyone and really are sorry for everyone that's affected. Reliability is actually a lot better than it used to be across everything, but the only problems people see now are the really big ones, because everything else we have enough redundancy that we cover it up More seriously, we take the failed machine out of circulation within a few seconds and only the users who were connected at that exact instant notice anything. Having extra hardware has made that easier.

[disneylogic]

... As speedy a resolution as ever!

Despite other opinions here, I'll stick with FastMail as My Main, despite having accounts of record elsewhere.

Works well, feels like home, have tons of stuff here, trust it.

Why would I change?

[veloster]

Originally Posted by brong

It's one particular type of new hardware that has now had a second failure in exactly the same way, and we have further debugging information this time that we can take back to the vendor and ask for more information.

I'm not a network infrastructure expert, but if a single piece of "new hardware" can take down not only all your POP and IMAP access but even your webmail and website, that seems like a design problem--especially if it's happened before.

I'm used to redundant connections to the net, redundant routers, redundant firewalls, redundant load balancing, etc. It's usually the guy trying to run everything out of a single rack at some co-located site (or in his basement) that has critical single points of failure such as you're describing.

In addition to sending debugging information to the hardware vendor, perhaps you need to re-visit the architecture so one piece of hardware can't create such a massive outage?

[digp]

I think you are asking for too much for too little.

Perhaps try https://luxsci.com/extranet/email-hosting.html#overview

[brong]

Originally Posted by veloster

I'm not a network infrastructure expert, but if a single piece of "new hardware" can take down not only all your POP and IMAP access but even your webmail and website, that seems like a design problem--especially if it's happened before.

I'm used to redundant connections to the net, redundant routers, redundant firewalls, redundant load balancing, etc. It's usually the guy trying to run everything out of a single rack at some co-located site (or in his basement) that has critical single points of failure such as you're describing.

In addition to sending debugging information to the hardware vendor, perhaps you need to re-visit the architecture so one piece of hardware can't create such a massive outage?

We have redundant everything, but it's not redundant against one piece of networking hardware doing something really awful at the low level. We are indeed sending debugging information to the vendor, and looking into how we can consolidate the damage to a single enclosure - probably by upgrading the trunks between our internal switches so that a single faulty host can't swamp the entire network. It's basically a DOS attack that's going down, but caused by a piece of hardware on our network! It has the bandwidth to swamp out everything else.

[veloster]

Originally Posted by digp

I think you are asking for too much for too little.

Perhaps try https://luxsci.com/extranet/email-hosting.html#overview

Thanks for the link. I also use Gmail for free, Microsoft Live mail for free, and Yahoo mail for free. Subjectively at least, all 3 have been more reliable than Fastmail. Granted they have large companies behind them, but they're also even less expensive than Fastmail.

I realize no service is perfect, everything is a trade off, etc. But for a company that specializes in just one thing--email--it doesn't seem reasonable to have such a critical single point of failure.

[digp]

Subjectively? Precisely.