I'm not aware of any permanent Roman settlement on the Canaries. The islands had been known about since the Carthaginian navigator Hanno, and apparently there was some trade during Late Antiquity, but I would be surprised at anything more substantial.

For colonies in general, it depends on what you mean. In the Roman sense, which requires a substantial resettled population, the answer is no, the Roman settlements were always within their borders. Even in a modern sense that does not require substantial settlement Rome did not posses colonies as distinct from the rest of its empire. It held effective suzerainty over certain client states, but I think it would be a mistake to equate this with colonization. I also don't really think that the relation of Rome to its provinces is very analogous to that of the modern colonial powers with their colonies, except perhaps at the very early stages. England and Scotland is a rather closer comparison than Britain and Nigeria.

All that being said, there were almost certainly Romans resident in areas outside the borders. The Indian city of Muziris seems to have had a population of Roman merchants significant enough to be mentioned in Tamil poetry, a Roman era merchant manual eludes to Roman merchants resident in Barygaza (modern Bharuch), and it is possible there were also merchants in cities like Mantai (in Sri Lanka) and Arikamedu. Outside of India it is rather more difficult to know as there hasn't been as much research on Roman trade along the West Coast of Africa, but it is possible that there would also have been some resident at Adulis (in Eritrea). There is reason to believe that by the middle second century CE, individual Romans had traveled around Mainland SE Asia, but how wee are supposed to interpret that is a matter of debate.

It is worth keeping in mind that this is not a system of dominance as you see in the colonial period, and just as there are Roman merchants in India, classical literary sources mention Indian and African merchants in Alexandria and a Tamil graffito has been found in the Red Sea port of Berenike. Individual regions might be controlled by a political power which controlled distribution of the resources from there, the most typical example being Moscha Limen/Khor Rori (modern Oman) which was a major producer of frankincense, the export of which was tightly controlled by the Hadramite king, but no individual group of merchants would monopolize a particular trade politically or legally. Even though the Persians were the primary export partner of the Hadramite king, he had no qualms with selling to Romans or Nabateans, and there is no evidence that any group attempted to apply political pressure to get a more favorable trade position.

This is not to say that political players were entirely uninterested in the trade, a well known illustration being Aelius Gallus' invasion of Arabia in the time of Augustus, but these were focused on controlling production rather than distribution. Two potential wrinkles in this is the recent discovery of a detachment of Roman soldiers stationed on the Farasan Islands off Saudi Arabia, and the Hadramite control of the island of Socotra. Unfortunately we really do not know what to make of these.

Hands down the best starter work for all things about Rome's Eastern trade is Gary K. Young's Rome's Eastern Trade. It is focused on history rather than archaeology, but is definitely the place to start. If you are lucky and you have access to Topoi Orient-Occident there is a very readable article by Fauconnier in the 2012 supplement ("Autour du Périple de la mer Érythrée") that deals with this, and don't worry, it is in English.