3 ms·

this is a recommended setting even on Oracle [1] [2] and Couchbase [3] the doc recommends adding "transparent_hugepage=never" to the kernel boot line in the "/

this is a recommended setting even on Oracle [1] [2] and Couchbase [3] the doc recommends adding "transparent_hugepage=never" to the kernel boot line in the "/etc/grub.conf" file. [1] http://oracle-base.com/articles/linux/configuring-huge-pages-for-oracle-on-linux-64.php#disabling-transparent-hugepages http://oracle-base.com/articles/linux/configuring-huge-pages... [2] https://support.oracle.com/epmos/faces/DocContentDisplay?id=1557478.1 https://support.oracle.com/epmos/faces/DocContentDisplay?id=... [3] http://blog.couchbase.com/often-overlooked-linux-os-tweaks http://blog.couchbase.com/often-overlooked-linux-os-tweaks

antirez 12y ago
Yes, never reported by Redis users so far, maybe since in Redis it does not have the same impact it has in other databases, but a more subtle latency issue apparently. EDIT: Actually in the CouchBase case they talk about "page allocation delays" that looks related, potentially. I started the investigation mainly because the Stripe graph looked suspicious with a modern EC2 instance type that has good fork times, so there was something more indeed (unless the test was performed with tens of GB of data).