11 ms·

I am so scared to imagine what would happen if there was any issue during the move (very likely when dragging live cables and powers over hundreds of meters).

I am so scared to imagine what would happen if there was any issue during the move (very likely when dragging live cables and powers over hundreds of meters). The client would immediately refuse to pay anything because he was very clear he wouldn't pay a thing if there is downtime. Then, the next contractor would be super quick to judge you and the situation, reinforcing that you were an incompetent idiot and the client was right to kick you away on the spot and not pay a dime. Glad it went well in the end. There is so much to lose for the person trying to help.

willcipriano 6y ago
This is a junior sysadmin I suspect. With a bit more experience you'd learn to say something along the lines of "no downtime, sure, that will be 30 grand" and the ability for downtime will suddenly materialize. Him and his friend did this big song and dance, took a huge risk and only got paid for ten hours worth of work in the end.
- zrail 6y ago
  $30k, 80% up front, strict liability waiver that says I’m not responsible for loss of business or anything else if there is downtime.
  - pedrocr 6y ago
    Now you're over-charging massively. If you have no liability and are guaranteed pay, charging for just double hourly rate is more than enough as a "stupid and non-standard requirements" kind of thing.
  - imtringued 6y ago
    You can't get paid upfront and at the same time get a liability waiver. For a 100% guarantee with full liability $30k doesn't actually sound ridiculous because it would require obtaining 100% identical hardware and doing at least one test run on that hardware before actually doing it on the production hardware. What the contractor did is basically "wing it", explain a way to get zero downtime to the client and then not actually offer a guarantee by doing the operation straight on the production hardware. Really this was more about convincing (ie bullshitting your way through) the client to let you do the work than actually doing it properly and for a huge sum of money. It wouldn't surprise me if there was actual downtime for a few seconds and the client simply didn't notice it.
- cellularmitosis 6y ago
  > sure, that will be 30 grand I am having trouble finding a reference to it now, but I've heard patio11 refer to this as "the Japanese no". Don't ever say "no" directly, just quote an astronomical price.
  - x86_64Ubuntu 6y ago
    People in the trades world do it too. If a job won't provide the margin they are seeking, or the job is more difficult than it's worth they will up the price. If the consumer chooses them to do the job, it's at a pricepoint that's worth the trouble but they are really hoping to be passed over.
    - coldcode 6y ago
      The Rolling Stones tried that with Microsoft and "Start Me Up", they quoted what they thought was ridiculous $10M. Microsoft said sure, no problem.
      - Strom 6y ago
        That's a fun story. Looking more into it, it seems that $10M is based on rumors and it was more likely $3M. [1] Doesn't change the point of the story though. -- [1] https://www.networkworld.com/article/2220097/what-microsoft-paid-the-stones-to-help-launch-windows-95.html https://www.networkworld.com/article/2220097/what-microsoft-...
      - vntok 6y ago
        $10M is a debunked urban legend; the actual figure is only $3M, which is pretty standard. Microsoft's whole ad campaign for Win95 cost about $200M after all.
    - briankelly 6y ago
      My dad's in construction and frequently gives out "fuck off" quotes. It isn't so rare that the client accepts them.
  - vorpalhex 6y ago
ThePowerOfFuet 6y ago
Also, moving a server with spinning disks? What could possibly go wrong.
- daemin 6y ago
  Wasn't here a story about Sun (or HP or someone like that) where they moved a bunch of disk servers across a parking lot to another building and found that many of them had died from the vibrations on the trolley cart used to transport them.
  - znpy 6y ago
    it was Yahoo, IIRC.
- sschueller 6y ago
  I had a spinning disk in my car back before we had all these cool embedded PCs. The disk was never an issue, these things can take a lot of abuse (Even New England roads). I had it mounted sideways so a large pothole wouldn't push the heads into the platter.
  - perl4ever 6y ago
    >a spinning disk in my car https://www.consumerreports.org/cro/news/2014/04/record-players-were-the-infotainment-systems-of-the-1950s-and-60s/index.htm https://www.consumerreports.org/cro/news/2014/04/record-play... "The stylus did not jump the grooves even when the car was moving at various speeds over broken pavement, cobblestones, and deep holes."
- driverdan 6y ago
  Spinning disks can take a surprising amount of shock and vibration before they fail.
- nwallin 6y ago
  Disks aren't that sensitive to motion. At my last job, we had 2 airplanes with 5 computers each with 6 disks each mounted in an aircraft. These were regular servers from Dell, not special hardened or resilient hardware or anything. So 60 or so hard disks flying around. Takeoffs, landings, turbulence. Two flights per day, 3 hours each flight, 6 days per week. So 626 landings per year. Disk failures were not particularly common.
sleepybrett 6y ago
If there was downtime during the move and the client was there and declaring that they would not pay, you just walk away. You'd be surprised at how fast they can cut a check in that situation.
gnopgnip 6y ago
4.5 hours of a consultants billing rate can be much more than 10 hours of your regular hourly rate working a similar job. A good consultant will have a contract. The client saying I won't pay if XX happens doesn't mean anything unless it was in the contract. Networking/spanning tree loops, arp table mismatch/corruption, the switches at the destination being misconfigured are all realistic problems that would result in downtime here. The normal way you do this is with live migration from hyper-v or vmotion from ESXI. If the initial migration is not successful, you just leave the server powered on while you address the issues. Once the VM has been migrated you can do whatever you want with the original server without worrying about downtime.
- maire 6y ago
  This reminds me so much of when I joined vmware in 2006. vmotion had already been around for a few years - but I believe this was the first release of vCenter with DRS. A couple of months I joined, a room full of customers chewed us out for not publishing our vmotion compatibility tables. After 4 hours of chewing out - they then told us they reverse engineered the compatibility tables and reorganized their entire data center to conform to vmware vmotion. Then (of course) we worked with intel to make sure the compatibility matrix worked in the future. I realized at that point that I joined the right company.

willcipriano 6y ago

This is a junior sysadmin I suspect. With a bit more experience you'd learn to say something along the lines of "no downtime, sure, that will be 30 grand" and the ability for downtime will suddenly materialize. Him and his friend did this big song and dance, took a huge risk and only got paid for ten hours worth of work in the end.

zrail 6y ago
$30k, 80% up front, strict liability waiver that says I’m not responsible for loss of business or anything else if there is downtime.
- pedrocr 6y ago
  Now you're over-charging massively. If you have no liability and are guaranteed pay, charging for just double hourly rate is more than enough as a "stupid and non-standard requirements" kind of thing.
- imtringued 6y ago
  You can't get paid upfront and at the same time get a liability waiver. For a 100% guarantee with full liability $30k doesn't actually sound ridiculous because it would require obtaining 100% identical hardware and doing at least one test run on that hardware before actually doing it on the production hardware. What the contractor did is basically "wing it", explain a way to get zero downtime to the client and then not actually offer a guarantee by doing the operation straight on the production hardware. Really this was more about convincing (ie bullshitting your way through) the client to let you do the work than actually doing it properly and for a huge sum of money. It wouldn't surprise me if there was actual downtime for a few seconds and the client simply didn't notice it.
cellularmitosis 6y ago
> sure, that will be 30 grand I am having trouble finding a reference to it now, but I've heard patio11 refer to this as "the Japanese no". Don't ever say "no" directly, just quote an astronomical price.
- x86_64Ubuntu 6y ago
  People in the trades world do it too. If a job won't provide the margin they are seeking, or the job is more difficult than it's worth they will up the price. If the consumer chooses them to do the job, it's at a pricepoint that's worth the trouble but they are really hoping to be passed over.
  - coldcode 6y ago
    The Rolling Stones tried that with Microsoft and "Start Me Up", they quoted what they thought was ridiculous $10M. Microsoft said sure, no problem.
    - Strom 6y ago
      That's a fun story. Looking more into it, it seems that $10M is based on rumors and it was more likely $3M. [1] Doesn't change the point of the story though. -- [1] https://www.networkworld.com/article/2220097/what-microsoft-paid-the-stones-to-help-launch-windows-95.html https://www.networkworld.com/article/2220097/what-microsoft-...
    - vntok 6y ago
      $10M is a debunked urban legend; the actual figure is only $3M, which is pretty standard. Microsoft's whole ad campaign for Win95 cost about $200M after all.
  - briankelly 6y ago
    My dad's in construction and frequently gives out "fuck off" quotes. It isn't so rare that the client accepts them.
- vorpalhex 6y ago

ThePowerOfFuet 6y ago

Also, moving a server with spinning disks? What could possibly go wrong.

daemin 6y ago
Wasn't here a story about Sun (or HP or someone like that) where they moved a bunch of disk servers across a parking lot to another building and found that many of them had died from the vibrations on the trolley cart used to transport them.
- znpy 6y ago
  it was Yahoo, IIRC.
sschueller 6y ago
I had a spinning disk in my car back before we had all these cool embedded PCs. The disk was never an issue, these things can take a lot of abuse (Even New England roads). I had it mounted sideways so a large pothole wouldn't push the heads into the platter.
- perl4ever 6y ago
  >a spinning disk in my car https://www.consumerreports.org/cro/news/2014/04/record-players-were-the-infotainment-systems-of-the-1950s-and-60s/index.htm https://www.consumerreports.org/cro/news/2014/04/record-play... "The stylus did not jump the grooves even when the car was moving at various speeds over broken pavement, cobblestones, and deep holes."
driverdan 6y ago
Spinning disks can take a surprising amount of shock and vibration before they fail.
nwallin 6y ago
Disks aren't that sensitive to motion. At my last job, we had 2 airplanes with 5 computers each with 6 disks each mounted in an aircraft. These were regular servers from Dell, not special hardened or resilient hardware or anything. So 60 or so hard disks flying around. Takeoffs, landings, turbulence. Two flights per day, 3 hours each flight, 6 days per week. So 626 landings per year. Disk failures were not particularly common.

sleepybrett 6y ago

If there was downtime during the move and the client was there and declaring that they would not pay, you just walk away. You'd be surprised at how fast they can cut a check in that situation.

gnopgnip 6y ago

4.5 hours of a consultants billing rate can be much more than 10 hours of your regular hourly rate working a similar job. A good consultant will have a contract. The client saying I won't pay if XX happens doesn't mean anything unless it was in the contract. Networking/spanning tree loops, arp table mismatch/corruption, the switches at the destination being misconfigured are all realistic problems that would result in downtime here. The normal way you do this is with live migration from hyper-v or vmotion from ESXI. If the initial migration is not successful, you just leave the server powered on while you address the issues. Once the VM has been migrated you can do whatever you want with the original server without worrying about downtime.

maire 6y ago
This reminds me so much of when I joined vmware in 2006. vmotion had already been around for a few years - but I believe this was the first release of vCenter with DRS. A couple of months I joined, a room full of customers chewed us out for not publishing our vmotion compatibility tables. After 4 hours of chewing out - they then told us they reverse engineered the compatibility tables and reorganized their entire data center to conform to vmware vmotion. Then (of course) we worked with intel to make sure the compatibility matrix worked in the future. I realized at that point that I joined the right company.