Dissonance is in the ear of the beholder and is also context dependent. For example, Richard Strauss, in Ein Heldenleben, writes the semitonal cluster D / E flat / F flat, scores it for oboes and cor anglé doubling three trombones and marks it sfz and then repeats this twice; it is underpinned by a diminshed seventh chord whose root is A. Similarly, Chopin writes a chord including a semitonal cluster E# / F# / G underpinned by an F# towards the end of his Scherzo No. 1 in B minor and marks it fff and then repeats it eight times (measures 594-599 inclusive); curiously, he does the same in his later Polonaise-Fantaisie three measures before the Più Lento section in B major (same notes as in the Scherzo example and also underpinned by an F#), although here the dynamic is pp and there are no reiterations.
These three examples are all "dissonances" in one sense and yet they are each in a specifically tonal context which might be argued to mollify their "dissonant" impact.
Best,
Alistair