[HBASE-22300] SLB doesn't perform well with increase in number of regions

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Component/s: Balancer
• Labels: balancer
• Assignee: David Manning
• Reporter: Biju Nair
• Created: 23/Apr/19 20:55
• Updated: 10/Jul/21 22:57
• Resolved: 10/Jul/21 22:53

Description

With increase in number of regions in a cluster the number of steps taken by balancer in 30 sec (default balancer runtime) reduces noticeably. The following is the number of steps taken with by balancer with region loads set and running it without the loads being set i.e. cost functions using region loads are not fully exercised.

Nodes  regions  Tables    # of steps           # of steps 
                          with RS Load         With no load   
5       50       5        200000               200000
100     2000     110      104707               1000000                          
100     10000    40       19911                1000000                          
200     100000   400      870                  1000000                          

As one would expect the reduced number of steps also makes the balancer take long time to get to an optimal cost. Note that only 2 data points were used in the region load histogram while in practice 15 region load data points are remembered.

Issue Links

• is fixed by: HBASE-25832 Performance improvements for StochasticLoadBalancer