Normally when you type 'free' you get an output that shows:
total used free shared buff/cache
available
Mem: 9.3Gi 226Mi 5.5Gi 9.0Mi 3.6Gi 8.9Gi
Swap: 15Gi 0B 15Gi
Where total is the total amount of RAM actually installed in the
machine. In this case, this is only showing that 9.3Gi.
Checking with cat /proc/meminfo shows the same 9.3Gb of RAM as the Total.
However, checking with "xl info" it shows:
total_memory : 16356
Yesterday the 'free' command was showing that it only had 4 GB of RAM.
I rebooted the entire server and brought the 4 VMs back online and it
showed the total to be 16GB again.
Is this normal? Seems like a big memory leak. But I'm wondering if
it could be that I have some bad memory and it keeps mapping it out on
the fly.
I do notice that the VMs get slower and slower over time. Just
rebooting everything will clear it up, but it slows down over about 2-3
months time to the point where ssh to them takes 10-15 seconds to connect.
total used free shared buff/cache
available
Mem: 9.3Gi 226Mi 5.5Gi 9.0Mi 3.6Gi 8.9Gi
Swap: 15Gi 0B 15Gi
Where total is the total amount of RAM actually installed in the
machine. In this case, this is only showing that 9.3Gi.
Checking with cat /proc/meminfo shows the same 9.3Gb of RAM as the Total.
However, checking with "xl info" it shows:
total_memory : 16356
Yesterday the 'free' command was showing that it only had 4 GB of RAM.
I rebooted the entire server and brought the 4 VMs back online and it
showed the total to be 16GB again.
Is this normal? Seems like a big memory leak. But I'm wondering if
it could be that I have some bad memory and it keeps mapping it out on
the fly.
I do notice that the VMs get slower and slower over time. Just
rebooting everything will clear it up, but it slows down over about 2-3
months time to the point where ssh to them takes 10-15 seconds to connect.