Jump to content
  • Advertisement
Sign in to follow this  
SteveDeFacto

GameDev.net has the most down time of any site I regularly visit.

This topic is 2712 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.


Recommended Posts

Advertisement
Our provider is investigating what happened - definitely must have been a fluke type thing for there to be that level of corruption. We're on a LAMP stack now.

Share this post


Link to post
Share on other sites
Here is the full explanation from our provider of the events that transpired:


[font=arial, sans-serif]

I have reviewed ticket number 3059161 in full. From reading the ticket[/font]
[font=arial, sans-serif]

history, it does appear that Rob had taken care to ensure that your instance[/font]
[font=arial, sans-serif]

was returned to an up and running state as soon as possible. I would also like[/font]
[font=arial, sans-serif]

to assure you that the issues which took place in that ticket are not typical[/font]
[font=arial, sans-serif]

for our servers.[/font]

[font=arial, sans-serif]

However, the initial technician who handled your request to perform the fsck[/font]
[font=arial, sans-serif]

did in fact neglect to run a fresh backup of the server prior to doing this.[/font]
[font=arial, sans-serif]

This may have been due to there being a backup image from January 5th and he[/font]
[font=arial, sans-serif]

may have assumed you would not want to have a fresh backup image created with[/font]
[font=arial, sans-serif]

a corrupted filesystem. [/font][font=arial, sans-serif]

There is also the possibility that attempting to[/font]
[font=arial, sans-serif]

create a backup image prior to the fsck would have caused more harm than good.[/font]
[font=arial, sans-serif]

Likewise, it appears from the logs that the backup which was created on the[/font]
[font=arial, sans-serif]

6th failed due to the raid failur[/font][font=arial, sans-serif]

e. This then lead to loss of data on your end[/font]
[font=arial, sans-serif]

from the 6th and needing to restore from the 5th, where it may have been[/font]
[font=arial, sans-serif]

possible to salvage data from a backup had it been created prior to the fsck[/font]
[font=arial, sans-serif]

being run. In a situation such as this, it is always best to receive[/font]
[font=arial, sans-serif]

clarification though and I assure you this will be addressed and corrected.[/font]

[font=arial, sans-serif]

In regards to the redundancy which we have in place for our servers, we[/font]
[font=arial, sans-serif]

do have a raid 10 on the parent server for the partition where customer data[/font]
[font=arial, sans-serif]

is stored, which means a few disks could fail before causing major issues in[/font]
[font=arial, sans-serif]

the array. We also have our servers setup so that backups and images are[/font]
[font=arial, sans-serif]

stored on seperate servers in case there are failures with the hardware or the[/font]
[font=arial, sans-serif]

raid setup of the originating parent server.[/font]

[font=arial, sans-serif]

One particular issue that arose, which did not help the situation, occurred[/font]
[font=arial, sans-serif]

when your server was first restored to the new parent. It appears that the[/font]
[font=arial, sans-serif]

instance was still active on the old parent as well and network traffic was[/font]
[font=arial, sans-serif]

primarily being routed to the old parent. At this time, are you still[/font]
[font=arial, sans-serif]

experiencing further issues in regards to this? If so, I would be more than[/font]
[font=arial, sans-serif]

happy to get them resolved for you.[/font]

[/quote]

Share this post


Link to post
Share on other sites

Here is the full explanation from our provider of the events that transpired:


[font=arial, sans-serif]

I have reviewed ticket number 3059161 in full. From reading the ticket[/font]
[font=arial, sans-serif]

history, it does appear that Rob had taken care to ensure that your instance[/font]
[font=arial, sans-serif]

was returned to an up and running state as soon as possible. I would also like[/font]
[font=arial, sans-serif]

to assure you that the issues which took place in that ticket are not typical[/font]
[font=arial, sans-serif]

for our servers.[/font]

[font=arial, sans-serif]

However, the initial technician who handled your request to perform the fsck[/font]
[font=arial, sans-serif]

did in fact neglect to run a fresh backup of the server prior to doing this.[/font]
[font=arial, sans-serif]

This may have been due to there being a backup image from January 5th and he[/font]
[font=arial, sans-serif]

may have assumed you would not want to have a fresh backup image created with[/font]
[font=arial, sans-serif]

a corrupted filesystem. [/font][font=arial, sans-serif]

There is also the possibility that attempting to[/font]
[font=arial, sans-serif]

create a backup image prior to the fsck would have caused more harm than good.[/font]
[font=arial, sans-serif]

Likewise, it appears from the logs that the backup which was created on the[/font]
[font=arial, sans-serif]

6th failed due to the raid failur[/font][font=arial, sans-serif]

e. This then lead to loss of data on your end[/font]
[font=arial, sans-serif]

from the 6th and needing to restore from the 5th, where it may have been[/font]
[font=arial, sans-serif]

possible to salvage data from a backup had it been created prior to the fsck[/font]
[font=arial, sans-serif]

being run. In a situation such as this, it is always best to receive[/font]
[font=arial, sans-serif]

clarification though and I assure you this will be addressed and corrected.[/font]

[font=arial, sans-serif]

In regards to the redundancy which we have in place for our servers, we[/font]
[font=arial, sans-serif]

do have a raid 10 on the parent server for the partition where customer data[/font]
[font=arial, sans-serif]

is stored, which means a few disks could fail before causing major issues in[/font]
[font=arial, sans-serif]

the array. We also have our servers setup so that backups and images are[/font]
[font=arial, sans-serif]

stored on seperate servers in case there are failures with the hardware or the[/font]
[font=arial, sans-serif]

raid setup of the originating parent server.[/font]

[font=arial, sans-serif]

One particular issue that arose, which did not help the situation, occurred[/font]
[font=arial, sans-serif]

when your server was first restored to the new parent. It appears that the[/font]
[font=arial, sans-serif]

instance was still active on the old parent as well and network traffic was[/font]
[font=arial, sans-serif]

primarily being routed to the old parent. At this time, are you still[/font]
[font=arial, sans-serif]

experiencing further issues in regards to this? If so, I would be more than[/font]
[font=arial, sans-serif]

happy to get them resolved for you.[/font]


[/quote]

Yes, that does sound like a fluke but time is the only way to know for sure.

Share this post


Link to post
Share on other sites
I understand the point of this thread, but I can't shake the feeling that it is an unwarranted slap in the face to the creators of the site. Did you really question investing your time in GDNet? They've already given so much that I feel it is disrespectful to cavil about such a small matter. A matter that they are completely aware of and actively work on everyday.

If you can do it better than them, do it. If you are tired of this site, leave. Otherwise, find something more constructive to comment on.

Share this post


Link to post
Share on other sites
To be fair, just because you criticize or "complain" about a community, product, or idea that you have some concern for doesn't make you disrespectful. There's nothing wrong to expect quality service or question to the status quo. No one here is being ungrateful. Matter of fact, the fact that the forum is using a third party software and not a hand rolled one would lead one to believe that it will be pretty stable and have very little downtime. If users feel that the site is going down too often or for too long then they have every right to voice their concerns. The majority of the people posting have been here since before the upgrade. So they're not some new kid, know-it-all running their mouth.

With all that said, transitions are hard but the site has been good so far. Once this forum and its intricacies have been grokked, we'll all finally shut up about this and sing its praises.

Share this post


Link to post
Share on other sites

If users feel that the site is going down too often or for too long then they have every right to voice their concerns.

I do not dispute their right, but almost assuredly the moderators are already well aware of their concerns. Which is why I believe our attention could be turned to something more constructive.

To the same point, the original post simply struck me as a little too acerbic for my tastes. Generally I would not even comment on such a matter.

Share this post


Link to post
Share on other sites

I understand the point of this thread, but I can't shake the feeling that it is an unwarranted slap in the face to the creators of the site. Did you really question investing your time in GDNet? They've already given so much that I feel it is disrespectful to cavil about such a small matter. A matter that they are completely aware of and actively work on everyday.

If you can do it better than them, do it. If you are tired of this site, leave. Otherwise, find something more constructive to comment on.

Totally agree with this.

If you like the site, don't just support it with your visits and posts.

Although posting does add some value, it doesn't add money; every page view has a cost, however small.

You can pay to become a GD+ member (giving you some extra features and a gold name) and you can make a cash donation (giving you the "benefactor" award).

If you don't want to pay, at make sure your ad blockers are off on the sites you want to support, and please occasionally click through an add or two. You don't have to look at it, just click through it. Make it a personal quota "I'll follow 3 adds per day to support the site" or something.



Point is, it takes money to run a quality site. If you aren't giving them money you don't have much room to complain.

Share this post


Link to post
Share on other sites
Regardless of what GD.net does, it ought to toss a donation link:
[color=#0000ff]Donate | Watched Content | New Content

...at the top of every page. Even if it only generates a small amount of income, at least it'd generate something month by month.


Yes most downtime by a longshot, its always been thus though, even 10 years ago (ironic since its a computer focused website).

On the flip side, I've yet to see a site that has better SEO. =)
Post almost anything in a forum, and it'll be on top of google's results within an hour, even if those same terms are getting lots of attention elsewhere on the 'net by million+ user sites.

Share this post


Link to post
Share on other sites
If we could get [color=#0000ff]Donate | My Content | Watched Content | New Content I'd be even happier.

Yeah I know if I click on my name I can get to the link from there but..... honestly that's a bit of a pain.

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!