The possible formation of stellar-mass binary black holes through dynamical interactions in dense stellar environments predicts the existence of binaries with non-negligible eccentricity in the frequency band of ground-based gravitational wave detectors; the detection of binary black hole mergers with measurable orbital eccentricity would validate the existence of this formation channel. Waveform templates currently used in the matched-filter gravitational-wave searches of LIGO-Virgo data neglect effects of eccentricity which is expected to reduce their efficiency to detect eccentric binary black holes. Meanwhile, the sensitivity of coherent unmodeled gravitational-wave searches (with minimal assumptions about the signal model) have been shown to be largely unaffected by the presence of even sizable orbital eccentricity. In this paper, we compare the performance of two state-of-the-art search algorithms recently used by LIGO and Virgo to search for binary black holes in the second Observing Run (O2), quantifying their search sensitivity by injecting numerical-relativity simulations of inspiral-merger-ringdown eccentric waveforms into O2 LIGO data. Our results show that the matched-filter search PyCBC performs better than the unmodeled search cWB for the high chirp mass ($>20 M_{odot}$) and low eccentricity region ($e_{30 Hz} < 0.3$) of parameter space. For moderate eccentricities and low chirp mass, on the other hand, the unmodeled search is more sensitive than the modeled search.