i’ve just spent way too much time trying to figure out why the same query run on one VM with elastic 0.1s while on another – 2s. both VMs were running on the same physical server, had the same amount of cores allocated.
i’ve finally found it – there was a tiny difference in amount of RAM allocation. the faster one had:
<memory unit='KiB'>67108864</memory> <currentMemory unit='KiB'>67108864</currentMemory>
which is exactly 64 * 1024 * 1024 * 1024
slower one had:
<memory unit='KiB'>65536000</memory> <currentMemory unit='KiB'>65536000</currentMemory>
this should teach me not to type those numbers by hand