Comment by tooltalk
9 days ago
Some foolishly believed that the twin towers were invincible after the 1993 WTC bombing.
Before 9/11, most DR (disaster recovery) sites were in Jersey City, NJ just across the river from their main offices in WFC or WTC, or roughly 3-5 miles away. After 9/11, the financial industry adopted a 50+ miles rule.
Jersey City still was fine and 50 miles can be problematic for certain types of backup (failover) protocols. Regular tape backups would be fine but secondary databases can't be that far away (at least not at the time). I remember my boss at WFC saying that the most traffic over the data lines was in the middle of the night due to backups - not when everybody was in the office.
Companies big enough will lay the fibre. 50-100 miles of fibre isn't much if you are a billion dollar business. Even companies like BlackRock who had their own datacenters have since taken up Azure. 50 miles latency is negligible, even for databases.
The WTC attacks were in the 90s and early 00s and back then, 50 miles of latency was anything but negligible and Azure didn’t exist.
I know this because I was working on online systems back then.
I also vividly remember 9/11 and the days that followed. We had a satellite dish with multiple receivers (which wasn’t common back then) so had to run a 3rd party Linux box to descramble the single. We watch 24/7 global news on a crappy 5:4 CRT running Windows ME during the attack. Even in the UK, it was a somber and sobering experience.
12 replies →
In the US, dark fiber will run you around 100k / mile. Thats expensive for anyone even if they can afford it. I worked in HFT for 15 years and we had tons of it.
5 replies →
Before 9/11, most DR (disaster recovery) sites were in Jersey City, NJ just across the river from their main offices in WFC or WTC, or roughly 3-5 miles away. After 9/11, the financial industry adopted a 50+ miles rule.
IIRC, multiple IBM mainframes can be setup so they run and are administered as a single system for DR, but there are distance limits.
A Geographically-Dispersed Parallel Sysplex for z\OS mainframes, which IBM has been selling since the '90s, can have redundancy out to about 120 miles.
At a former employer, we used a datacenter in East Brunswick NJ that had mainframes in sysplex with partners in lower manhattan.
If you have to mirror synchronously the _maximum_ distances for other systems (e.g. storage mirroring with NetApp SnapMirror Synchronous, IBM PPRC, EMC SRDF/S) are all in this range.
But an important factor is, that performance will degrade with every microsecond latency added as the active node for the transaction will have to wait for the acknowledgement of the mirror node (~2*RTT). You can mirror synchronously that distance, but the question is if you can accept the impact.
That's not to say that one shouldn't create a replica in this case. If necessary, synchronize synchronous to a nearby DC and asynchrone to a remote one.
For sure we only know the sad consequences.
1 reply →
>Some foolishly believed that the twin towers were invincible after the 1993 WTC bombing.
I was told right after the bombing, by someone with a large engineering firm (Schlumberger or Bechtel), that the bombers could have brought the building down had they done it right.