-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 In the final, I do not think Squid architecture is designed for fast access to huge amounts of memory. It came from a time when computers were young, memory cost like Boeing and hardly Squid itself seriously reworked in this part since then. 23.10.2016 18:15, Yuri Voinov пишет: > > > > 23.10.2016 18:09, Matus UHLAR - fantomas пишет: > >> 23.10.2016 17:40, Yuri Voinov пишет: > >>> This effect is good known to all who have worked with relational > >>> databases. In fact, it is typical in general for all caches except > >>> purpose-built highly scalable systems. > > >>> 23.10.2016 17:37, Matus UHLAR - fantomas пишет: > >>>> doesn't that imply kind of effectiveness? > > > On 23.10.16 17:53, Yuri Voinov wrote: > >> In fact, the explanation is very simple. At some point soon will get the > >> content from the disc using an index of any kind than consequentially > >> and fully scan the giant structure in RAM. > >> > >> Performance indicator is expressed in the data access time. But this is > >> from the category of personal experience. Everyone can choose their own > >> road to hell. :) > > > saying this you could say that huge in-memory cache for OSes is > useless.... > This is not me talking, and tuning specialists. > > > iirc in some squid versions itwas caused by linear searching for memory > > objects. > Only up to a point and is highly dependent on the server hardware > architecture and software process architecture. > > > using indexes should speed up saerching and bigger probability to find and > > higher probability to find object in memory could outweight searching > time. > Certainly. As I said above, it depends on the software architecture > access to a cache memory. If there is an effective index structure (like > a balanced B-tree), the effect disappears. > > > databases are much faster when using indexes properly, aren't they? > Absolutely yes. But: Most database index structures exist for disk > objects and do not exist for the memory structures. There's a completely > different mechanism. Indices disk objects (which are the metadata) are > loaded into memory and used to access the on-disk data. But access to > the memory is carried out mostly through a simple list structures. > > Which leads to performance degradation in case of huge caches. Again we > come back to the importance of software architecture of the memory accesses. > > -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBCAAGBQJYDKrmAAoJENNXIZxhPexGMPEIALfRoGL7EDTi4lS1rmItes6k VJATrwChT8uZR+nexYTusVRaiYc2OnbQthTLYSOTYWqXl43l+Haj0+YAAc4edS1J 8ajAY0FzmGZsLynTsRLq526QsIXBUcuTAnbXbZb16g8sHbWZ/cnZzeR2SBP86qyC b7EqAsQV1preTlqXo4WpfaZFDZldjwTaemjb91Rl9HsdBaKxcru35wzZbdvefbng y4I9QPAv6xVFvEjw/IsUdpSe9vRPHFaLXF7WyYF2rM9+1mmCWd9YhFK0NASJ8Jxs REhMi/WMZHFCh3SK1Dj5H+bH4xONN0AHuf0YogpbgvmjxrJkQhRB88RkH2xNiRE= =/YWN -----END PGP SIGNATURE----- |
Attachment:
0x613DEC46.asc
Description: application/pgp-keys
_______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users