New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Mine always freezes for running 2-3 days continuously.
That's not right. Laptop shouldn't be doing that, even in a Windows environment.
doing what? stay up 3 months? or freezing 2-3 daily?
I'd rather say you don't have to clear swap at all. If something was swapped out, its there on purpose. Linux are mature enough to manage VM. At least with physical nodes.
But you can tweak /proc/sys/vm/swappiness if you want to change behavior.
I've had 45 days uptime on my MBP before, only rebooted for updates.
I had 60 days uptime on my old EeePC I've sold, had to reboot it because my sister needed to use Windows
Work computer running W7 Had 69d uptime (get on my level) then crashed 5 days ago when I plugged my phone. Heh.
You guys think too highly of Windows environments. 9/10 times at least one of your drivers or applications has issues with Windows' power management functions. Good luck coming out of sleep more than 5 times in a row on a Windows laptop without some sort of issue. Luckily for me, the issues are almost completely exclusive to CPU throttling and touch screen drivers.
My computer at home is almost Never shut down.
Rock solid stability, and that's with Windows 7.
My laptop with windows 8 well that's a different story. But that's why I'm getting a new one.
I completely agree, Linux does a great job at using swap only with good reason. So word to all, if your machine swaps, let it. It's not necessarily bad. In some app servers where the apps are entirely in RAM, an operator error (aka ME) may mis-manage the RAM and see a performance drop when apps are supposed to be fast in RAM but actually wound up in swap when memory got full, and in a production environment where performance is key, a 2 minute server reboot during non-peak hours can be the quickest short-term fix with minimal impact. Every environment is different, and not saying rebooting a Linux box is ever a good decision technical-wise, but I can see a business case where it is a good decision in a case like this. As an engineer, you want to ask why it's happening, and spend time troubleshooting when it happens. As a business person, you want it fixed now because it's affecting the bottom-line. There's always a balance.
@dedicados if your app runs best with a reboot every 100 days, do it. It's probably something wrong with your app, because it is very abnormal on Linux. To be sure though, check your syslog files for any system messages, your app/database logs for warnings or unusual behavior, check the bandwidth usage charts for perhaps an outside source causing the problem, and check that the system clock has the right time.
Thanks for the tip!
guess what!!
i restarted and server crash. xD
how can that be possible, if i dont install anything and i dont get also error from system
but anyway, now is fixed,
but with a raid 10 this could be avoided? or is ok just with raid1 ?? is for hosting with cpanel
sure, webhosting is more reading and barely writing except for MySQL
thanks bronze, what do you recommend, raid10 is faster?
Organized downtime for upgrades/testing- yes.
Randomly rebooting at 100 days - no.
Hm I did pass on this thread a while Ago and it's back so Just for the record the question is not whether rebooting But whether you can boot your server. Some of the first responses to this thread are just bad. If you can't boot the machine you have a bigger problem than whatever may be hogging resources. For production environments I always prefer a more default stack with more resources on the box over configuring and or rebuilding things. My experience is all. This is a grievous offense. But also one of the reasons I enjoy scouring the LEN for good deals. And as a systems admin it lets me focus on directing the entirety of web needs. There's a good thread here right now touching on the SQL needs for types of clients lots of you guys have. It was nice to see some one mention the often poor quality, however, of the queries themselves and applications behind them. Poor MySQL and the bad rap any RDMS can always suffer from poorly written code.
The problem isn't the actual rebooting of the machine, it's the downtime caused by rebooting.