And keep in mind that starving the OS of memory to give it to the JVM is an anti-pattern, see Uwe's excellent blog on MMapDirectory here:
http://blog.thetaphi.de/2012/07/use-lucenes-mmapdirectory-on-64bit.html Best, Erick On Wed, Jan 7, 2015 at 5:55 AM, Shawn Heisey <apa...@elyograg.org> wrote: > On 1/6/2015 1:10 PM, Abhishek Sharma wrote: >> *Q* - I am forced to set Java Xmx as high as 3.5g for my solr app.. If i >> keep this low, my CPU hits 100% and response time for indexing increases a >> lot.. And i have hit OOM Error as well when this value is low.. >> >> Is this too high? If so, how can I reduce this? >> >> *Machine Details* 4 G RAM, SSD >> >> *Solr App Details* (Standalone solr app, no shards) >> >> 1. num. of Solr Cores = 5 >> 2. Index Size - 2 g >> 3. num. of Search Hits per sec - 10 [*IMP* - All search queries have >> faceting..] >> 4. num. of times Re-Indexing per hour per core - 10 (it may happen at >> the same time at a moment for all the 5 cores) >> 5. Query Result Cache, Document cache and Filter Cache are all default >> size - 4 kb. >> >> *top* stats - >> >> VIRT RES SHR S %CPU %MEM >> 6446600 3.478g 18308 S 11.3 94.6 >> >> *iotop* stats >> >> DISK READ DISK WRITE SWAPIN IO> >> 0-1200 K/s 0-100 K/s 0 0-5% > > Your questions cannot be easily answered. We can make guesses, but in > the end, figuring out how much hardware and exactly what configs to use > is something that only you can determine, by actually trying it: > > https://lucidworks.com/blog/sizing-hardware-in-the-abstract-why-we-dont-have-a-definitive-answer/ > > The following URL is the only general guideline I know of, and because > of the problems mentioned on the blog post above, it's not all that > helpful for specifics. Full disclosure of my bias ... I wrote most of > this wiki page: > > http://wiki.apache.org/solr/SolrPerformanceProblems > > Any recommendation we make will err on the side of caution, and may > involve spending more money for your hardware than you intended. I > personally would not try to get a Solr install going on machines with > only 4GB of RAM unless it was a VERY small index. Your mentioned heap > size of 3.5GB is quite small compared to what we normally see here. My > own production heaps for Solr are 6GB. > > Thanks, > Shawn >
