Jump to content

Server Downed


Ashoat

Recommended Posts

I thought Helio has been closed down without any reason, like those noob hosts I used in the past.

Me Too. I was hoping to use HelioHost for a clan site I'm currently making, But when I found both my personal site and the HelioHost home offline I thought HelioHost was gone for good!

 

Also my site void.heliohost.org isn't back online for some reason. It doesn't really mater (I mainly used it for testing and stuffing around) but I'm just letting everyone know that not all the sites are back online.

Link to comment
Share on other sites

  • Replies 65
  • Created
  • Last Reply

Top Posters In This Topic

It still seems to be crashing daily.

 

I agree with Wizard.

The crash has been decreasing since few days ago.

Link to comment
Share on other sites

The issues are still occurring. After lots and lots of hours of debugging, I have concluded that the current server crashes as well as the "hacking" that happened earlier were hardware issues. To be specific, it seems like Stevie is having memory issues. Memory issues are really hard to detect and often result in really bizarre problems (which I thought must have been a hacking last time). I am currently working with our colocation provider to see if we can confirm these issues, after which point I will request a replacement RAM stick under Stevie's warranty.

 

Here is a summarization of the issues I have experienced so far:

- About three weeks ago Apache went down on Stevie. When I attempted to restart it, Apache informed me that its port was in use. When I tried to check which application was using the port using netstat –p, lsof, fuser, etc., my SSH session just froze. The shell connection was still there, but the server was unresponsive. I was able to restart these frozen sessions, however, by reconnecting. After playing around, I found that anything involving a PID, or access to the /home partition would crash in this fashion. I concluded that somebody had hacked Stevie, and requested that he be shipped from the datacenter to my house so I could reinstall the kernel.

- After Stevie arrived, I started him up to see if the issue could have been specific just to OpenSSH. I was surprised to find that not only were the issues completely gone when directly logged in, but they did not appear when connected over SSH. After strengthening the server’s security a little bit and checking for rootkits through an external kernel, I shipped the server back. He seemed to work fine.

- However, soon after the server came back online, I started to notice more issues. Apache was generating large amounts of coredumps (about one every four minutes). After investigating the coredumps, I found that httpd -t (a configtest, equivalent to apachectl configtest) would generate a segfault once in about twenty or so times. What’s more, it would segfault in completely random places; there was no consistent pattern. I now conclude that my server has memory issues.

Link to comment
Share on other sites

So is that means if you can locate the problems with Stevie, and then request hardware replacement under warranty, everything will be fixed?

Link to comment
Share on other sites

Well, scratch the memory idea. It's not a memory issue.

 

I have absolutely no idea when this issue will be fixed. I'm working on it as much as I can, but I can't guarantee anything.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.

×
×
  • Create New...