Member Avatar for Dani

I've gone and done a whole bunch of refactoring today, specifically related to query string parameters and page navigation, and so please let me know if you encounter anything out of sorts. Also, sorry about the downtime earlier in the day. All part of coding in production :)

  • 5 Contributors
  • 8 Replies
  • 62 Views
  • 5 Days Discussion Span
  • Latest Post Latest Post by Dani

Recommended Answers

All 8 Replies

Member Avatar for Dani

So my IDE is set up that if I save a file, it automatically uploads it to production. That's how bad I am. However, at least I only gave my IDE read-only database permissions.

Member Avatar for Salem

So my IDE is set up that if I save a file, it automatically uploads it to production. That's how bad I am. However, at least I only gave my IDE read-only database permissions.

Yeah, you're just one fat-fingered moment from disaster.

Do you not have a test system replicated from the live DW at some regular interval to at least test with?

Member Avatar for Pak_052

It sounds like a successful day! There are always some bumps in the road when refactoring, so thank you for keeping things going and for the heads-up!

Member Avatar for Reverend Jim

Our AGC/SCADA system (pre 1998) was a dual system with an A and B cpu (one online, one standby). Each side had its own hard drive (300 meg) partitioned A and B. The online system ran from the A partition of the online cpu and the backup from the B partition of the backup cpu. All development was done in the B partition of the backup CPU. We could take the backup down and boot it into a pseudo-online mode with limited hardware to test database and software changes. Once everything was verified we would halt the online and fail over into the B partition, running everything live from B. We had a lengthy checklist to verify functionality. Once verified we would copy all the changes into the A partition of that cpu then restart and test again.

At this point the former online cpu/drive were still offline. If testing went OK then we'd remove the old online drive and put in a new drive. Once everything was copied from the new online drive to the old one, the old cpu was booted as backup.

More disk copies to make backups and a ton of saves to mag tape.

When the entire provincial electrical grid (transmission and distribution) depends on a stable and available system strict guidelines for testing andf integration are a must.

We maintained better than 99.99% availablilty over a twenty year period.

Member Avatar for Dani

We have multiple web servers in a load balancer, as well as a master/slave database setup. Years ago we had 6 web servers, but as traffic declined over the years, we couldn't afford them anymore. Only 2 were purchased when we switched to colo. However, we can only live with one of the web servers taken down for very short periods of time, and they each handle different background tasks as well. Additionally, we only have one email spooling server, one search server, etc.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.