Table 8 Comparison of various migration methods along with TPO method
From: A three phase optimization method for precopy based VM live migration
S.no. | Migration techniques | References | Parameters | Overhead | Workload | ||
|---|---|---|---|---|---|---|---|
TPT | TMT | DT | |||||
1. | Compression based | Ma et al. (2012) | 50.5 % \(\downarrow\) | 48.2 % \(\downarrow\) | 47.6 % \(\downarrow\) | — | Idle, static web, dynamic web, stream video, kernel compiler |
Svärd et al. (2011) | 51 % \(\downarrow\) | 35 % \(\downarrow\) | Factor 10–20 \(\downarrow\) | — | LMBench benchmark, streaming video server | ||
Zhang et al. (2010) | 56.6 % \(\downarrow\) | 34.9 % \(\downarrow\) | 23.16 % \(\downarrow\) | — | Compilation, VOD, static web, dynamic web | ||
Jin et al. (2009) | 68.8 % \(\downarrow\) | 32 % \(\downarrow\) | 27.1 % \(\downarrow\) | — | Static web, dynamic web, kernel-compiler, dbench, MUMmer (memory-intensive) | ||
2. | CPU scheduling based | Jin et al. (2011) | — | 1s \(\uparrow\) in TMT | 88 % \(\downarrow\) | 6 % More CPU usage | Static web, dynamic web |
3. | Shared storage based | Jo et al. (2013) | — | 30 % \(\downarrow\) | — | 0.2 % Space | RDesk, admin, file I/O |
4. | Prediction of frequently updated pages based | Alamdari and Zamanifar (2012) | 68 % \(\downarrow\) | 62.3 % \(\downarrow\) | 0.4 % \(\downarrow\) | — | Idle, kernel built, memtester 4.3 (memory-intensive), stress 1.0.1 |
Ma et al. (2010) | 34 % \(\downarrow\) | 32.5 % \(\downarrow\) | \(\uparrow\)* | — | MUMmer (memory-intensive) | ||
Liu et al. (2011) | Reduced migration cost 72.9 % in terms of MT and DT* | — | Linux idle, TPC-C, Dbench, LINPACK, SPECweb2005 | ||||
Hu et al. (2011) | \(\uparrow\)* Bound no. of iterations 3–5 | \(\uparrow\)* | \(\uparrow\)* | — | Low and high dirty page environment | ||
TPO | 71.1 % \(\downarrow\) | 70.4 % \(\downarrow\) | 3.4 % \(\downarrow\) | 0.05 % Space | Idle, kernel built, memtester 4.3 (memory-intensive), stress 1.0.1 | ||