Server not responding

Relating to problems with the DedicatedServer / MasterServer / DataCenter
User avatar
strategy-(DOG)-
Full Game Admin
Full Game Admin
Posts: 1666
Joined: Wed Jul 27, 2011 5:06 pm
Location: Austria
Contact:

Re: Server not responding

Post by strategy-(DOG)- »

....back with good news! :D

actually, the support found errors on the secondary harddrive. meanwhile this drive has been replaced.
the dedicated server is online again and currently in production mode synchronizing the harddrives (as long as this procedure will run, the server might cause lagging).

...the tension subsides ;)

WOOF!
[ no tolerance for intolerance ]
User avatar
x21
-(DOG)- Council Admin
-(DOG)- Council Admin
Posts: 163
Joined: Sun Jun 08, 2014 12:35 pm
Location: 地̸̛̤̥̒͆獄̵̌͜ͅ

Re: Server not responding

Post by x21 »

Have they only replaced the one drive or have they replaced both?
User avatar
strategy-(DOG)-
Full Game Admin
Full Game Admin
Posts: 1666
Joined: Wed Jul 27, 2011 5:06 pm
Location: Austria
Contact:

Re: Server not responding

Post by strategy-(DOG)- »

there were no errors or issues on the primary drive. only the secondary drive has been replaced.
meanwhile the harddrives are synchronized.
[ no tolerance for intolerance ]
User avatar
Dr.Flay
FOD
FOD
Posts: 89
Joined: Wed Mar 26, 2014 10:36 am
Location: Cornwall
Contact:

Re: Server not responding

Post by Dr.Flay »

This illustrates my complaint of many years, that people do not use SMART to prevent downtime, but use it during downtime to diagnose how broken it is.

2 common problems;
1) No OS is setup to use SMART fully, and will only inform you of disk errors when the disk is failing, and SMART is giving-up on the drive.
2) Most drives do not have all 3 SMART options enabled by default, and people never check (have you on your PC ?).
These 3 switches are usually; 1) auto self test. 2) make a log. 3) keep previous logs.
If the ability to keep logs is not enabled, you have no history of the drive failure info.

The best use of SMART for a server (or any PC) is to predict failure based on constant monitoring.
Watching for spikes in temperature, or acceleration of the block retirement can all be good indicators before your drives actually fail.
With monitoring software you can set alerts to be sent when certain thresholds are passed.

Personally I would request that my servers are being SMART monitored proactively not retroactively, and what the status is of the 3 SMART settings on each of the drives.
Post Reply