Forum archive
CPU Utilization and Reserved Micro Instance
I am using a micro instance for VTiger, for the first 3 weeks it worked fine. Easy to access and backup. It is accessed about 5 times a day and only for a short time and not much processing. And now for the last 3 days it is extremely slow and and back up takes ages and I had to stop it. Amount of data to back up is small. Only a few mbytes. What is happening? When I looked at the CPU utilization graph on server dashboard it shows that utilization for the last 3 days is 100%. How come the utilization is 100% if nobody is accessing it. Is it the cpu utilization of my instance when nobody is accesing it? If it is, then how come for the first 3 weeks the cpu utilization is not 100%. I am not a technical guy I don't understand it.
If I change it to a reserved micro instance, will it solve the problem? Any help is appreciated
Replies (5)
Been there
Reply 1I posted recently regarding a similar experience
Reply 2So from what I know Chris is totally on the money (in fact I'm sure he knows much more about AWS from a practical perspective than me).
See the recent post I was referring to here.
Micro reserved
Reply 3Unfortunately the answer is no.
Micro reserved in my opinion are not really worthwhile.
You may get a very cheap vps but when it holds 100 percent cpu for longer than aws threshold you get pushed aside and you are not responsive.
From my long term tests, 18 months now, i see micro go 100 perce t often.
From what i can tell when an update coincides with a backup or a stuck process, ther are lots of reasons.
I tried rolling my own tinycore which uses hardly anything and it never failed.
No slight on the devs but tkl images are just a bit too much for any semi serious usage (edit - on micro servers the others are good) Some will disagree but i have hard proof.
I have used them for low volume websites and they are ok. Start playing with scripts that do fancy things and it gets ugly.
The minute i tried small instances, i was happy.
Temporary Solution
Reply 4First of all, thanks to both of you. I also posted the problem via the blue tab on the server dashboard, and Alon suggested that it might be the issue related with the instance or application. Thanks to Alon for his prompt response.
I read somewhere in the forum that email might cause a problem and VTiger has an email module which is enabled even though I do not use it. So I disabled the email module and reboot the instance and now for the last two days the instance is back to normal again, meaning it is running well like the first 3 weeks. Let me see if it is o.k for another few weeks. If it does not work out I will try Chris suggestion to go to small instance. Hopefully it works.
Chris is right
Reply 5Hi Chris, you have been there and you are right. Microinstance is only good for hours but days later CPU utilisation went up 100% and my instance was pushed aside. Now I have upgraded to a small instance and so far it is working all right.
Typically Micro instances are designed for "spikey" CPU utilisation.
If they consume 100% for any length of time they will be degraded, possibly why 100% is shown.
Another factor I have found is that of server utilisation.
Micro instances seem to be more or less a way for AWS to sell off additional resources on a server that is otherwise not crunching data. When they do crunch data you are kinda pushed aside. When you stop and restart your server you will notice its all good again. This can be explained because its now on a less utilised server. I have provisioned a small instance and every issue I had with micro went away.
I just had a quick google and found THIS. Looks like I was on the right track :)