They just announced a $5 hike in price this month for "fanatical" support, I haven't opened a support request with them in eons:
> In July 2014, Rackspace introduced a minimum support fee of $50 for all new accounts to offset the rising cost of providing Fanatical Support. Legacy accounts like yours were unaffected by this change, keeping your previous rate without a support fee for the past four years. Effective February 2019, you will see the addition of a $5 support fee on your invoice.
I've been grooming a elastic IP @ AWS for the last year, and rackspace is getting dropped this week.
A well behaved remote mail server should either reject/bounce the message, or deliver it to a user's spam folder, it should never just drop it. Unfortunately, It's near impossible to determine if your message ended up in a user's spam folder :(
My server sends 50-100 messages a month, tops, and I never send mass messages, just direct messages to family, friends & customers. The biggest issue I had was when Rackspace periodically (+ voluntarily) reported their VPS IP space to Spamhaus, which required me to inform Spamhaus that I was responsible for that specific IP, and it should not be on the blacklist.
Is this new ? I have been part of discussions on HN about personal mail servers for ... 8 years ? This is the first I have heard someone mention registering their mail server with google ...
You don't have to register per-se, but you get access to reports for your domain so you can see any potential trouble.
Corporate email admins can also be opinionated about email senders because their email system is only for their employees' work, specifically. They are less likely than Gmail to care about email from the little guys who are trying to build an audience or business. They definitely don't see it as their job to provide startups and random people with access to the inboxes of their staff. They will err on the side of saying no to unfamiliar email.
Corporate email is where it gets hard. You can have a clean IP, all the authentication options configured correctly, and still struggle to get out of the spam folder. This is one reason businesses will pay for big-name email platforms, which have spent a lot of time and money building good reputations with corporate teams, and have people who specialize in maintaining those reputations.
Disclaimer: I'm one of the developers behind RBLMon.
Also, as a customer I do not mind thinking about a business' costs. In fact I want to know that a business is pricing a reasonable margin above costs and no more. I have way more respect for companies that don't see the need to mindlessly pad their profits. Not everyone is you.
(Not saying RS is being honest here but it is important to push back against this "pay no attention to what's going on behind the curtains" thinking)
How is their fanatical support lately? I recall my company (at the time) paying through the nose for a pretty responsive Fanatical Support SLA, such that I could call in at 2AM and almost every time get someone in no more than three rings-central time zone, no less-for our dedicated services account.
Curious how it is for other customers who might not have had a tech budget the size of ours and could swing heavy hammers to get such support.
About your mailservers: location doesn't matter, as reputation is being build over the time. I have Digital Ocean $5/1GB machine that was previously used for heavy spamming (had to submit tickets with all biggest SpamHouses) and since then has pristine reputation with constant 97-99 SpamScore, etc. You just need about 45 days of constant raising volume of mailouts and you be fine. This machine currently does about 40,000 mailouts a day.
DO's equivalent Elastic IP product can't be used to originate email, unfortunately. They actively block it, and it's against their ToS (Last time I looked). Otherwise I would have preferred to host it there. If you allocate a Droplet, groom it's IP, and then terminate it (either for yours or their reasons), they only "promise" to give you back the same IP, and I'd hate to have to setup a new IP in a rush.
Before I setup my DO I asked them can I set up ESP? They said absolutely. AFAIK they only ban constant abusers. I sent to very clean and engaged list (noone above 45 days of inactivity) and hadn't had any problems. I have "emergency setup" on different hosting but you are right on point it would take month to come back to same mailing levels.
DO TOS doesn't say you can't send emails. Their control panel allows you to setup rDNS yourself BTW.
These days largest ESPs (Google, Yahoo/Rocketmail, AOL/Live/Outlook/Hotmail, GMX, Wanddo, Orange, UOl, BOL) only really look at IP in 4 categories: narrow whitelist (other biggest mailers, private contracts [banks mostly]), wide whitelist (static IPs of all sort including your $5 DO), greylist (home IPs, mobile IPs, all other IPs that probably should be mailout out in the first place), blacklist (in terms of you land there forever - don't confuse with having your IP blacklisted). Long gone times when ESPs tried to figure on some scale 1-100 how good your IP is.
I've been using mailinabox and am generally happy with it as is mostly a hobby server, but am curious what others are running.
I very much miss when their support truly was fanatical.
Edit: duration of service, clarity
For example, we agreed a planned maintenance window 2 weeks ago. It happened, it failed and they left the server offline for 7 hours after the window cut off.
We asked of compensation was available. It took 7 days just for the account manager to say no. I can accept that it's in the ToS but is 7 days for a response from our account manager 'fanatical support'?
We're in the process of moving our WP sites to WPEngine, where the support is much better, and software we maintain for clients to AWS for a fraction of the cost either way.
I could make a call and be instantly connected to a Linux tech that was often a better sysadmin than anyone we had on staff, which was why we really encouraged clients to pay the premium for their service. The choices then were a) host cheap and pay $150-200K in sysadmin salaries to keep it going while passing on these costs to clients or b) host at RS and use their economy of scale. Easy decision.
Then they got bought by Apollo and I hoped what I"d seen happen with other companies bought by private equity wouldn't happen. The PE model only gives a shit about profit, which is at odds with a service focus. Rather than seeing profits as the by product of a good model, excellent service, and loyal customers, profits become the measuring stick by which operational costs are measured.
So that Linux tech that cost $120K+ per year gets binned and replaced with two juniors who cost less but together are unable to accomplish in thrice the time what the binned guy could do. Hold times start to creep up because shit isn't getting fixed. So they hire more juniors who still can't get things done quickly.
Guys like me put up with this for a brief time and then bail because the pricing can no longer be justified.
I'm sure it's true that the bulk of their cost is support, but it's not fanatical anymore because the folks that really were so have been gone for years.
That's not the case anymore and it's clear nobody gives a hoot. We are getting a DDoS a few months ago. Their solution: null route our server. Wanted DDoS mitigation options. They wanted me to speak to a sales rep, who insisted it wasn't a sales call, and refused to put their possible options in writing. As a rule, I only make decisions or open discussions if and only if I have something to discuss in writing. That never used to be a problem. Everything was transparent. It's become a nightmare driven by sales reps that want to "customise" a solution - the same kind of solution that they deployed a few years ago that worked for the exact same problem.
Bingo. Back in the day the person I reached on that first phone call really did give a shit _and_ they had the skillset combined with the authority to find and implement solves.
At the end of our working with them whether the person I spoke with cared or not wasn't the issue any more: it had taken 30 minutes plus to get to them and it was a very junior person who didn't have the skills or authority to run an ls -la from the bash prompt.
I'm still pretty bitter, obviously! My company ate a lot of costs moving clients away from RS. And in many cases we had to rebuild our credibility a bit after the company we glowingly recommended started failing to hold up their end.
The culture started going downhill around the time they were bringing Bank Of America employees into cloud operations upper management and they were chain hiring their pals.. Lot of talent was disease franchised and leaving.
Never had any downtime and the only degradation I can remember was when they got hit by a collosal DDoS a few Christmases ago (since then they've invested very heavily on ensuring that can't happen again, US DC's where down/sporadically up for days).
Hands down my default choice.
Polishing off a resume doesn't mean immediately quitting either- it means going out to look at other opportunities. There are very few scenarios where being aware of other options is harmful.
Supposedly this is an annual thing now: https://www.tpr.org/post/third-year-row-february-rackspace-l...
Typical slog of working at a recently PE purchased company.
I wonder if anyone talented is left. Usually they get sick of it after a year or two. I would guess little to no real RSU handcuffs. The stock was badly underwater before it went private.
Except we hadn't completed ours yet, this event was planned for several months due to leadership expecting double-digit turnover after the performance review period, about 10% of leadership in a single product got wind of it coming and has already settled in at new companies, multiple principals and architects with spotless records were let go from a single product along with nearly 80% of its other headcount, and layoffs have been quietly happening without attention being drawn (a few a week). We were also told at an Open Book that Apollo had to intervene for Rackspace to afford contracted bonus structure this year, as Rackspace literally did not have cash on hand to meet targets.
Other than all that, though, astute insight and analysis. You're right, just typical fiscal year slog. (PR is literally saying "we're a stable company, promise" in statements.)
Especially when the people who had been given assurances by ELT and then woke up Thursday to find they'd been broken on account of near insolvency of the business, when those people are hanging out in the comments among you, it's probably best to listen more than speak.
I was guessing review/stack/rank at the beginning of the year because I've spent time working at recently PE bought company, and quit once I saw the pattern.
When do managers write reviews? They don't have to be delivered to the employees before rank/stack/layoff...
My manager was terminated before he could complete reviews for his staff (he was forbidden from telling any of us about it, because they wanted us to be surprised we were being fired, so my immediate manager just up and vanished one morning). His manager, and that person's manager both turned over from the company weeks ago after they were given a courtesy notice of this event coming. Rackspace then reacted to the turnover and forced the highest manager mentioned so far to stay and personally fire everyone.
I understand that you think your theory regarding this layoff is correct. It is not. Performance reviews had absolutely nothing to do with this event. At all. This event involved cost center.
Proof I'm legit: The meeting invite to fire everyone was called "Business Update," and scheduled by Liana Greenberg on behalf of Scott Crenshaw. They fired 30 at a time, then quickly cleared the room for the next batch.
Big hosts can survive by essentially offloading support, and making as much as possible self-managed by the end-user.
The non-technical people who just want "a server that works" won't, broadly speaking, pay a premium for that. They're non-technical and will eat up support-hours too, asking questions about setting file permissions on Wordpress.
If everyone just charged some basic $ for support, it would generally get better. I'm perplexed and sometimes shocked how much 'free support' people expect - and often "get" to some degree - by cheap hosts. I'm in some WP support forums on facebook and... man - what people expect for $10/month is crazy. And... the amount of bs misinformation and half-truths being spread by 'experts' in those forums in crazy too.
Add in 'authoritative' sites and experts 'recommending' (via affiliate links) bluehost/godaddy/siteground on a recurring basis, and we've got this... again(?) - race to the bottom?
People will post how 'awesome' the support is because some remote people spent 15 minutes walking them through file permission setup amazes me.
Oh... and "free migrations" between hosts. Then when there's a real issue - like 6 of the 38 plugins don't actually work on the new host, and the new host says "this will be $75 to make it all work" people are outraged! and publicly slam everywhere they can. Then someone else promises unlimited XYZ, they move, and the cycle starts again.
Finally... liar/cheats/scumbags who intentionally rip people off on top of all that (I think some of the ripping off isn't always intentional, just misinformed) makes people even more wary of trusting genuine people.
I may have just described half of the history of human society, unrelated to hosting, but.... I've watch this market grow over the past 20+ years, and it's just so... weird.
A lemon market is when the buyer doesn't have enough information to tell the difference between a good product (peach) and a bad one (lemon).
Since the buyer can't tell the difference, the price he is willing to pay will end up somewhere between the value of the peach (more expensive) and the value of the lemon (cheaper). What happens then is that the sellers of lemons thrive while the the sellers of peaches end up leaving the market. This lowers the average price and creates a cycle where the price decreases which drive even more sellers of quality products to leave the market which lowers the price further...
Economies of scale https://en.wikipedia.org/wiki/Economies_of_scale
A Gresham's Law dynamic.
Rackspace might have had a chance, but it is gone now..
That is dead and buried now though. There may still be some good people there but their products are dated 'also ran' knockoffs of AWS services and I don't trust them within an inch of our environments anymore.
I felt bad for clients that had been suckered with that cost burden, particularly if the customers did not show up after go-live. It was like the online equivalent of them being lumbered with a giant mall when they needed a prominent spot on the High Street.
Nowadays though services like AWS have taken on this agency provided hosting setup.
You're also clearly unaware that Rackspace built itself and went public on colocation, which dates your perspective. "What type of servers to use" is such a variable question for each outfit, and even with a web-focused startup perspective like this one, there's so much variance in work that colocation often made sense at the time. Time sharing a VPS with Bob's IRC Vhost Service often does not.
The parent comment makes a valid point about performance-to-price value. DigitalOcean, Linode, and Vultr currently offer much more value for the newbie client (and for many other types of customers) than Rackspace does at the same price point.
Yet despite having a good idea of the requirements the project manager who has never coded in his life always gets clients a Rackspace box with enterprise Redhat and some horrid cPanel on it with some extra firewall box. These things hamper the go-live as if you come from Ubuntu then Red Hat always feels like it is a generation out of date with things turned off by default when you need them on. Then there is a cPanel sea of icons when all you need is an ssh terminal and 'apt-get'.
Some friction goes on because, as a developer you wonder how it is that you are not being listened to. Then one day it all becomes clear. There is this matter of a kick back which is a few hundred which is not really worth it for the hassle caused. If you are on the side of the client and you know full too well that they don't need to be paying out a grand a month for the few dozen sales they are getting in a week then you can't help but feel negative about Rackspace. The sales person at Rackspace has time and money to wine and dine your clueless middle manager whipping boy, but can't advise on a more sensible solution for your clients.
Eventually the clients move on to another agency because your agency have only thought of the 'go live' and no support beyond that, so in turn, Rackspace lose another customer once the expensive project has been torn to pieces and rebuilt by a better agency who charge the client for support instead of useless hosting that they will never get to need.
I am sure this is a scenario that was quite common a few years ago. Nowadays not so. Apart from anything else small businesses can just setup a Shopify website and be online by teatime rather than be bamboozled by useless web agencies.
"Average Monthly Recurring Revenue (MRR) of the first 3 full months' paid invoice"
And what someone like Linode pays:
"You then receive a $20.00 credit applied to your account"
The comment you're replying to is saying Rackspace gets some benefit from unethical agencies advising clients that don't need a dedicated server to get one anyway.
It wasn't a knock on Rackspace per se, more of a commentary on agencies.
My take is that they were pivoting HARD away from the server business and trying to provide "cloud architecture solutions", where they would tell you how to properly set up your GCP and Amazon clouds to fit your needs.
From inside the company, coming from the silicon valley, the culture was rather jarring. They're from San Antonio, and try to hard to be hip and silicon valley-ish, but it's just in the uncanny valley where everything is a little bit... off. Also, it might be the corporate culture in San Antonio, but they try really hard to incoporate the new employees into "the family". For instance, all new employees have to get flown to San Antonio for a week of team building exercises and orientation. To me, it almost seemed culty.
This is actually very SV way.
For example, this is not my first job out of college. I had to sit through a 2 hour meeting where a HR drone went ON and ON about tradeoffs of one health plan over another based on your life situation. Me, having basic reading comprehension, was on my laptop. I was reprimanded for that.
Cloud architects are hard to come by, and hard to pry away from their existing jobs.
Think about whether or not your want to pay Accenture marketing rates to get a junior employee who just started, because knowing what I know about the pay scale, if anybody gets some expertise in Cloud Architecture and has some contacts in tech, which is quite likely because he's talking to clients, he's probably going to jump ship for a company that pays better.
Rest In Piss.
Why not just retrain existing workers instead of laying them off? Stuff like this seems like the quickest way to kill morale and run the business into the ground.
But, also consider that the PE firm doesn't care all that much. They just need to make the numbers look good, so they replace highly paid veterans for cheaper junior workers. And ultimately, they force the company into treading water with a smaller/cheaper workforce which makes the numbers look good in the short term so they can sell it. Sucks for the company though.
Hands down expensive as hell but AMAZING support.
For Linux... Inmotion Hosting
Reasonably priced, but UNBELIEVABLE support.
Seems like everyone working at these two companies is a wizard that put on their robe and hat before work. When you pay for hosting you want someone on call and smarter than you for when things go south.
Their infrastructure feels like using a couple of half disassembled servers on the floor. Their platform is fragmented to pieces, they have no idea what a "design system" is, and you have to log into like three different webapps to just play around with your server.
They are literally in the same place they were in on day one, playform wise.
If you're nostalgic for some 2006 era hosting company, go with InMotion.
Back then Invotion/HDC was kind of shitshow, so I left and spent a couple years at Rackspace...
Pantheon for most Drupal sites. Amazing service. Should have made the move years ago.
Platform.sh for some Drupal sites and for various NodeJS projects.
Various CDNs for asset serving.
Almost everything is available now "as-a-service". The case where a custom server that you manage via SSH is the best solution is now the minority.
For years, Rackspace provided amazing sysadmin support, but there was always a gap between where their support ended, and the best practices for WP and Drupal sites began. That included tuning the LAMP stack for those applications, configuring dev sites and deployment processes, and managing caching and TLS. Rackspace helped with general advice, but we had to "do" it, and maintain it.
That's not true on CMS-specific hosting like WP Engine and Pantheon; it's all in one package: production hosting with scaling, dev and staging sites with one-click sync and deployments, CDN, TLS certs via Let's Encrypt, and support. We went from spending almost an FTE's worth of time on sysadmin and dev ops, to almost no time on those things.
I'd say I wish we had done it earlier, but in my opinion, we could not have. It has only been within the last year or two that these services have really put together the full feature set we wanted.
Unfortunately from reading some of the comments it appears that private equity took over RS at some point and now it's just gonna be their cash cow until they go bankrupt.
They haven't done much innovation wise in years. Its really sad.
I think the issue you don’t see “pop up” competitors is the inability to market “great support” empirically. There is a lemon market comment somewhere that is dead on. You can have all the high cost experts on staff but Everyone is going to think your marketing is full of it until they try your services or are referred via word of mouth.
I'm not sure there's any place you can get that sort of service now, which is part of the reason I run my own servers and dug more in to sysadmin stuff. Yes, my stuff's in a remote data center, and I can't physically touch it. Assuming there's a dumb pipe coming in, I can handle everything else (or find someone who can) - I don't need handholding service bundled in to a price.
And... they made a mistake early on. We'd requested a new machine, but someone had done the ticket wrong, and took the old one offline for an upgrade (instead of just adding a new machine to the account). They were minutes away from wiping it when I called - if I'd had an automated attendant, we would have lost everything on there, but humans intervened to stop them in time and get it back online. "Great service!" - except they'd made the mistake in the first place...
Rackspace support ca. 2004 was really fanatical. Like you could call them and get an engineer on the line, working on e.g. your RAID card, in minutes. Without going through a ticketing system or other intermediaries. Some of this was no doubt due to them being a smaller company, but "Fanatical Support" was not just marketing hype.
So far they have helped me with:
- S3 permissions loading files to RedShift and Aurora
- binary file support with API Gateway and lambda with CloudFormation
- a cross account Code Pipeline with CodeBuild, Code Deploy and CloudFormation - ended up you can’t do it fromthr vonfold.
- some problem a coworker was having with DAX (DynamoDB cache) with Python.
buy a lot of stock on number one (google, or Amazon in this case), then use your money or previous position on number two or three company in the same market to influence the board to drive the company value to the ground. now you can claim loss on the secondary position while creating a bigger monopolistic market for your main position for the long run.