Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On 19/12/12 01:19, Prater, James K. wrote: > Hello Andreas, > > Thanks for responding, below is a section of my current sysctl.conf file. As you can > see that I used the alternative parameters that you had sent me. The resource spikes > vary from 18-24 hours apart. The odd thing is that it sometimes occurs when the system > is just idling i.e. 5-18Mbytes a second. > > > # > # Kernel tweaks to close the gap between the buffercache and DRBD DIO to remote store. > # > vm.swappiness = 10 > vm.zone_reclaim_mode = 1 > vm.dirty_background_ratio = 1 > vm.dirty_ratio = 1 > vm.dirty_expire_centisecs = 300 > vm.dirty_writeback_centisecs = 60 > vm.vfs_cache_pressure = 40 > > The settings forces the delta between data coming into the file server and data being set to the peer to be > 1-3 seconds as opposed to nearly 25seconds. > Shouldn't these two values be different: vm.dirty_background_ratio = 1 vm.dirty_ratio = 1 ie, vm.dirty_ratio should be higher than vm.dirty_background_ratio. If they are both the same value, it means you never get the advantage of slowly catching up with your writes. This might be intended if you expect very consistent high write load... However, if you expect short spikes of high write load, then the vm.dirty_ratio could be higher, which would allow the system to catch up over the following low load period. Making them the same means that as soon as you have 1% dirty, you will stall all writes, and limit them to the speed of the underlying disk sub-system, which means you lose any advantage of buffering etc... Just my interpretation of the various meanings... Regards, Adam -- Adam Goryachev Website Managers www.websitemanagers.com.au