The Gear Page is down...

[Valhalla]

Last night around midnight got a time out.
A few minutes ago, a bad gateway message.
Hope they're back soon.
Edit - just got a 503 Service Temporarily Unavailable message

 

[Lef T]

Thanks for the update. I wonder what could have caused something like that...?

Sometimes gear goes down.
I've had two Marshalls crap out on me in my lifetime.
Maybe the Gibson buys Mesa thread short circuited everything.

 

[Ed Storer]

I just checked again at 11:07 PST and got
502 Bad Gateway502
It's a bummer for me because I just got off a 10-day suspension for skating on too thin ice - politics.
I like the site because it usually has more posts than TDPRI and Strat-Talk combined.

 

[getbent]

I just checked again at 11:07 PST and got
502 Bad Gateway502
It's a bummer for me because I just got off a 10-day suspension for skating on too thin ice - politics.
I like the site because it usually has more posts than TDPRI and Strat-Talk combined.

so, buffets are a good thing, right?

 

[RobDaglish]

I'm no expert, but I'm surprised they're still using physical servers and not a cloud-based solution (which I believe would be cheaper and offer more redundancy).

I’m in the field, but I’m more of the nuts and bolts then the sales end - but at the end of the day, the cloud is only someone else’s computer... and you’d be surprised, coatings on cloud can really stack up depending on what you’re doing to the point where you are better off buying 90K worth of computer and renting space in a data centre - I’m busy prepping some servers as a Proof Of Concept for a healthcare provider who are being stuffed with huge Azure cloud bills.

For a downtime this long, if feels like something serious happened. I can have clients switched to our DR site in minutes, restore a snapshot on our clusters in around 20 minutes, and have a new machine up from nothing in around two hours, then restoring data, probably a few hours on top of that. Hopefully once they’ve beaten it, we’ll get a nice Post Mortem.

It is typical to do software updates early Sunday mornings

When I worked at a school internet provider, they went for Network at Risk on Tuesday morning 4AM to 6AM. Seemed an odd time to do it, but that’s when we had the least traffic over the network!

 

[budglo]

I just checked again at 11:07 PST and got
502 Bad Gateway502
It's a bummer for me because I just got off a 10-day suspension for skating on too thin ice - politics.
I like the site because it usually has more posts than TDPRI and Strat-Talk combined.

I like it and like this forum as well. I used to go to MLP, but it’s pretty dead now like many gear related forums.

 

[archetype]

Interesting opinions from people re: physical servers, cloud hosting, etc.

Single instance, shared hosting, virtual machines, cloud structure: every bit of it runs on physical servers.

If people automatically say "cloud" is the answer to outages, remember that Google Cloud hosting has had "catastrophic" failures that took down YouTube, Gmail, Snapchat, Vimeo, Shopify, and thousands of other Web sites and services that powered businesses. The failures of Amazon Web Services have taken down Netflix, Amazon Alexa, Slack, Twilio, Quora, Reddit, and thousands of others.

There are reasons for using major cloud services and reasons for using smaller services that know your name. Both are subject to global issues and both are subject to user error by the service itself. Both can fail.

Yes, I've had enterprise-scale, managerial experience with Web hosting, computing, and data centers. I'm thrilled to be out of that business and no longer sleeping with a pager under my pillow and a cell phone by the bed.

 

[archetype]

I’m in the field, but I’m more of the nuts and bolts then the sales end - but at the end of the day, the cloud is only someone else’s computer... and you’d be surprised, coatings on cloud can really stack up depending on what you’re doing to the point where you are better off buying 90K worth of computer and renting space in a data centre - I’m busy prepping some servers as a Proof Of Concept for a healthcare provider who are being stuffed with huge Azure cloud bills.

For a downtime this long, if feels like something serious happened. I can have clients switched to our DR site in minutes, restore a snapshot on our clusters in around 20 minutes, and have a new machine up from nothing in around two hours, then restoring data, probably a few hours on top of that. Hopefully once they’ve beaten it, we’ll get a nice Post Mortem.
!

LOL! I just retired from project management in heathcare IS and my last gig was building out a data center for a hospital system and moving hardware and services between data centers.

Major cloud hosting was considered, but the security wasn't convincing and no one outside of healthcare understands the requirements of healthcare. Healthcare computing and storage aren't just generic services provided by a generic service.

You obviously get it.

Yeah, something blew up and there was no contingency. They're likely rebuilding something. Their Twitter page says the ISP is working a ticket for it, whatever it is.

 

[Swampash&Tweed]

This just went up....Technical issue at the server level.....Whatever thats means

 

[fjblair]

It's back up now.

 

[Rocky058]

All fixed



(image removed)

 

[DADGAD]

Funniest thing I ever read was the Pterodactyl thread on TGP. Someone actually posted that they saw one. Hilarious and epic. It actually helped me through a difficult time. Alas, it got deleted. Anyone remember it?

 

[ZackyDog]

It works again. Server was down.

https://www.thegearpage.net/board/index.php

 

[darkwaters]

Regular updates, including security updates, usually occur on the weekend. If you're on a site on the weekend and it's glitchy, chances are some lucky IT folks are working overtime to keep us amused and safe.

Sometimes it a minor set updates, no sweat; some are complex and everything goes well; sometimes... BOOM.

 

[archetype]

Regular updates, including security updates, usually occur on the weekend. If you're on a site on the weekend and it's glitchy, chances are some lucky IT folks are working overtime to keep us amused and safe.

Sometimes it a minor set updates, no sweat; some are complex and everything goes well; sometimes... BOOM.

Quite right, it can work out that way, but..

If it goes BOOM it's a sign of amateurs at work.

You don't update the production servers in production. You update the backup servers and, after they are tested and running correctly, you point the production address at the backup servers which makes them the production servers.

No down time and no one even notices that it happens except for a few folks that may need to refresh their browsers.

When it's all humming along nicely, you update the former production servers that are now the backup servers, then test them.

 

[Steerforth]

This is why I have my PC, iPad, and iPhone all connected directly into the Akashic Field. No downtime, no maintenance, no updates, and it has everything.