Tag Info

Hot answers tagged

5

ESXi doesn't support hot-removing memory or CPUs from a VM. Someone I met in the pub a couple of months ago told me that the sister of a friend of his cousin *cough* got the answer that virtually no customer demands this feature. In other words: It doesn't pay to bother implementing hot-remove. If you want ESXi to support this, go to VMware and tell them. ...


5

While it is true that kernel uses swap even if there may be memory left, using more than two thirds of it may be an indicator that historically the server may have been running out of memory and that is why it started swapping. I would correlate swap usage with memory usage in the sar reports to deduce whether the system has enough ram. I would also check ...


4

The Linux Kernel starts to swap out memory pages even if you have plenty of ram free. You can fine tune this behaviour by setting a custom swappiness. http://en.wikipedia.org/wiki/Swappiness For servers, I'd recommend to set the swappiness to 1 if you have always enough memory for your workload. For workstations, I'd recommend using the default of 60. # ...


3

The biggest thing to watch are page faults. On most Linux systems, running ps -o min_flt,maj_flt will give you some cumulative statistics, but real-time isn't always good enough. SAR is probably your biggest friend when it comes to checking out overall health of the system, including memory, processor, network, etc... Check out the man page for sar for ...


3

Why wouldn't you just correct the order? There's no reason to "what if" when you KNOW that this is not a recommended configuration. As for whether it will work, you will see once the server completes its POST. But I don't think it's necessary to play that game. Do you want to do this the right way? If so, spec and arrange the RAM according to the DIMM ...


2

I can not say for sure in vmware, but when using kvm decreasing memory of a running virtual machine works fine. Of course up to a limit, but the software will tell you. When you decrease it the memory in the virtual machine will gradually be lowered, until it can't be lowered any further. If you set the memory to be lower than where it stopped you will need ...


2

You should use the RAM slots adjacent to the CPU1 socket.


2

I hope it will be helpful to you. try performance_schema = off https://mariadb.com/blog/starting-mysql-low-memory-virtual-machines


2

I had the same problem with debian wheezy 7.8 guests. Installing the wheezy backports kernel 3.16.0-0.bpo.4-amd64 solved it for me. This was on the guests, I didn't touch the host. Add the following line to /etc/apt/sources.list: deb http://ftp.uk.debian.org/debian/ wheezy-backports main Then run apt-get update apt-get -t wheezy-backports install ...


2

TL;DR - Don't bump your memory limits for the sake of it. For a standard PHP application like WordPress, 128MB is huge and should be enough for 90% of your needs. However, there are a few important facts here: PHP Memory Allocation is per-script/process As per the PHP documentation for the max_memory option: This sets the maximum amount of memory in ...


1

As Craig Watson already told, you should not mess with PHP memory limit if not really necessary. Anyway, it is not possible to correctly size the server without a profound understanding of your workload. I suggest you to simulate a realistic load and in the meanwhile to monitor your server resource allocation/usage. For what it matters, 512 MB are very few ...


1

Row 3 : in top command indicates how the cpu is used. Understand each one separately % of the CPU for user processes (2.9%us) % of the CPU for system processes (2.3%sy) % of the CPU processes with priority upgrade nice (0.0%ni) % of the CPU not used (94.7%id) % of the CPU processes waiting for I/O operations(0.0%wa) % of the CPU serving hardware interrupts ...


1

You dont mention your distro/kernel version, but it may be worth noting that recent kernels (2.6.38+) have a feature called transparent hugepages. You can tell if it is enabled using: cat /sys/kernel/mm/transparent_hugepage/enabled It pretty much takes all this burden managing and preallocating hugepages of you. You can see how much memory is ...


1

Memory compatibility is always a toss-up, unless you get the OK from someone who's tried the exact same memory in the exact same system. That being said, I haven't heard of a case where inserting incompatible memory caused the magic blue smoke to escape from a system. At worst, it won't POST or you get an unstable system. So, try the off-brand memory and ...


1

No, it is just a live storage replication. Replication occurs asynchronously, periodically. Automatic failover would be created via script. See http://blogs.technet.com/b/keithmayer/archive/2012/10/05/automate-disaster-recovery-plan-with-windows-server-2012-hyper-v-replica-and-powershell-3-0.aspx as an example. Typically replica is used for DR, and ...


1

You may need to increase your Pagefile size, to be able to handle intermittent spikes in the memory commit size. We have this issue often in Azure compute where Pagefile is set WAY too low by default for memory-intensive apps. You can read more here: http://mvolo.com/low-pagefile-can-cause-503-service-unavailable-on-azure-web-roles/ This will not solve the ...


1

Are you experiencing any site performance issues? The thing with Linux is that it's not a bad thing for top to show that all your RAM is in use. Try giving us the output of free -m. When an application is finished with the memory it used, the kernel doesn't immediately purge the pages of data and mark it as free memory again. For instance, my desktop VM I ...



Only top voted, non community-wiki answers of a minimum length are eligible