CRYSTAL


 Cloud analytical databases employ a disaggregated storage model, where the elastic compute layer accesses data persisted on remote cloud storage in block-oriented columnar formats. Given the high latency and low bandwidth to remote storage and the limited size of fast local storage, caching data at the compute node is important and has resulted in a renewed interest in caching for analytics. Today, each DBMS builds its own caching solution, usually based on file-or block-level LRU. In this paper, we advocate a new architecture of a smart cache storage system called
 Crystal
 , that is co-located with compute. Crystal's clients are DBMS-specific "data sources" with push-down predicates. Similar in spirit to a DBMS, Crystal incorporates query processing and optimization components focusing on efficient caching and serving of single-table hyper-rectangles called regions. Results show that Crystal, with a small DBMS-specific data source connector, can significantly improve query latencies on unmodified Spark and Greenplum while also saving on bandwidth from remote storage.


affects the binding energies.
The free energy of formation of a mole of vacancies on the (100) surface of argon is :found to be liP = 1236 -3 .58 T cal This implies that at the melting point of argon there is one vacancy 12 ,. Argon is most suited for the calculation of defect properties. It is well known that argon is reasonably well represented by a Lennard-Janes 6-12 potential 4 wb.ich yield~ adequate energy values.by the direct summation of the bvo body potential function. Acco~dingly, we have chosen to_ examine certain surface properties . of argon. It is hoped that information gained from this simple solid will be of value in understanding the surface properties of more complex materials. -

UCRL-17161
Previously, the distortion of the perfect (100) surface of argon 5 6 has been calculated. ' It has also been shmm that the configuration of the perfect surface layer is probably identical 1·ri th that of the bulk except for the displacement of the surface myers e.1-ray from the bulk.. 7 A number of authors. have made a theoretical investigation of the configuration and energy of internal defects in argon.8,9 We have calculated the binding energies of argon and impurity atoms (neon and krypton) above and in the (100) surface plane of argon. We consider only the potential energy; therefore the results are reasonably v~lid only at 0°K. We also neglect all many-body forces. Vsing the high temperature Einstein approximation to calculate the ·entropy, "t·re estimate the concentration of vacancies in the equilibrium (100) surface of argon at its melting point assuming that the vacancies obey Boltzmann statistics. Though use of our data at the melting point is not justified by our assumptions, it is felt that the results obtained in this way are at least a ~ough estimate of the surface vacancy concentration at the melting point.
, We find, as ~xpected, that the binding energy to the argon t surface decreases in the series krypton, argon and neon. We also find, in accord with expectations, that the binding .energy of neon _and krypton above the perfect surface is less than that in. the surface plane. 'He find that thoUgh the relaxations are usually numeric.ally small, they appreciably affect the energy.
,.  Table l. It is also assumed that the concentration of defects is sufficiently small that defect-defect interactions are'neglectable.
Of these ass~mptions, the one most open ~o question is that of pairwise additivity of the potential function. For a perfect lattice this assumption is not seriously in error. Bullough, Clyde and Venables, 10 considering stacking faults in argon, have concluded that many-body forces contribute no more than .4% of the total binding energy. However,

11
. Sparnaay has estimated that an error of .as much as l0-3o% might be .made if van der Wal's forces are treated as pair-wise· additive; 12 Jansen has sho~~ that three body forces can explain the observed stability of the face-centered cubic form of argon over the hexagonal close-packed form.
Rossi and Danon 1 ) have indicated that the inclusion of three-body forces introduce a large error in the predicted energy of vaporization and attribute this error to either. four-body forces or a poor potential function.
We.have also used a Lennard-Jones 6-12 potential function to represent the interactions of neon and krypton atoms with argon atoms.
The argon-impurity potentials vrere obtained from the 6-12 potentials of ~' 4 Ne, 4 and Kr 14 in the following way. If rA-A represents the gas equilibri~m distance of atoms of type A and. UA-A is the depth of the vrell in the gas Subject to these assumptions, one ~ay write an expression for the total energy of the system as a function of the positions of all the atoms. One may then minimize the total potential energy of the system as a function of these positions and arrive at the energy and configuration of the relaxed defect. In practice this is not feasible, and we I .
assume that atoms not close to the defect are not dlsplaced from their normal positions. Only displacements of atoms close to the defective site are considered; we further assume that the distortions around the defect preserve the symmetry of the lattice, as much as is possible.
This last assumption was found to be warranted in the case of an internal vacancy9 and "' lve have not checked it for the surface problems. These  increase their separat:Lon from the relaxed nearest neighbors. This behavior is similar to that noted p~eviously for the relaxation of the atom~ ~round an internal defect.9 With a neon atom above the surface the relaxations of the nearest neighbors are smaller ( .2% dO'<Imrards and ·.2% ouhrards) while for krypton they are larger (.8% downwards and .4% o~twards).
It 1s interesting to note that in the fully relaxed situation, an extra neon atom lies closer to the bulk than an extra argon which, in turn, lies closer than an extra krypton atom.
With a vacancy in the surface, the nearest neighbors to the defect are. displaced towards the center of the vacancy (Fig. 4). Krypton           This report was prepared as an account of Government sponsored work.
Neither the United States, nor the Com-m1ss1on, nor any person acting on behalf of the Commission: A.
Makes any warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this report.
As used in the above, "person acting on behalf of the Commission" includes any employee or contractor of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, any information pursuant to his employment or contract with the Commission, or his employment with such contractor.