Okay a bit of economic theory to start us off. In any economy, people working for the government (e.g soldiers, school teachers, clerks) still need things like food and clothing and shelter. There's roughly three ways to supply said people with such necessities:

1. Direct requisition. Take a share of farmers' crops and animals, fishers' catch, demand the local town clothe and billet your troops - or directly levy labour, e.g. require serfs to spend X days a year working your fields, or the local lords must supply so many skilled knights for military service or the like. Obviously the general populace has less food, clothing, housing as a direct, short-term result of requisition. 

2. Taxes: demand money from the general populace.  Pay government workers with this money, the workers then buy their own food, clothes, shelter, etc. In the short-term the general populace buys less food, clothes etc than they would if they still had the money.

3. Inflation: mint/print money and give your new money to the government workers. They take the money to the shops/landlords and buy what they want. When the general populace comes along to buy their supplies, the shops/etc run out of their goods at the old prices. Thus prices go up, directly reducing demand and thus consumption in the short-term. (And indirectly encouraging more production, to some extent, but I won't try to summarise that debate right now.) 

(Of course in all these cases, the consumption of the general populace may well be higher in the long term because the soldiers stopped the Vikings or the teachers improved their students' productivity, I'm just speaking here of the short term.)

Each of these methods has pros and cons. For example it is a lot easier to move a chest of gold coins across the country (or mail a bank cheque) than move a thousand bushels of grain or 100 live chickens. On the other hand, the same properties that make money easier to transport than live chickens also make money easier to hide from the tax collectors. So we see a mix of the different funding mechanisms being used both in the same country over time and, often, in the same country at the same time for different situations. 

So in a market economy, money is not just a medium of exchange, and a medium of account (the book-keeping you mention), money is also a store of value. Taxing or inflation moves real resources from the general populace to the government's purposes similarly to how direct requisition does. 

Soviet Union

So how does this all apply to the Soviet Union, which wasn't a market economy?

For a start the Soviet Union needed real resources to fund its activities, from Soviet school teachers to Sputnik. The Soviet government couldn't just print 200 million more rubbles to build Sputnik, it needed metal and fuel and food to feed the workers and so forth. So the Soviet government faced budget constraints, in real terms, not just in terms of money. 

Beyond this, the Soviet economy was that it was, like all real-world economies, a mixed one. There were lots of deviations from state ownership, some legal, some illegal but widely tolerated by the authorities, and some illegal and persecuted, but still happening. Before the Russian Revolution, no one had had to really work through how "collective ownership of the means of production" would work, so unsurprisingly the Soviets discovered a lot of complexities and incentive problems, which they dealt with in varying ways.    Given incentive problems, the Soviet Union tolerated varying levels of private economic activity to varying amounts at different times.  A big legal category of non-government production was in agriculture. Collectivisation of agriculture in the USSR meant not so much state control, as more of a "cooperative organisation" - land in principle belonged to the government but was alloted rent free to the collective, along with small plots allocated to individual households. Livestock, in 1937, was partially owned collectively by the farm members and partly privately (Bergon, 1950, page 211).  Collective farms had to deliver some of their output to the state (paid at a very low price), but once they had met that and other obligations, they could sell their produce on the collective farm market. There was also a small prportion of independent peasants. 

Another category was "cooperatives", particularly in retail trade and small-scale industry, which weren't state owned. Reportedly (Hessler, 1998), a number of cooperatives tolerated their members operating privately and selling their services. There were self-employed - writers, musicians, also some artisans working for themselves, and clergy (Holzman, page 101). Also, to some extent, there were professionals like dentists with private practices, though my sources are obscure on how legal they were, or how extensive, Holzman (1958, page 100) for example just states that they faced higher taxes. Private sector activities were much more tolerated at times of economic stress, e.g. while the Soviet Union was struggling to recover from the Civil War (Lenin formally introduced the New Economic Policy in 1924) and during WWII, when state production was devastated by the loss of land to the Nazis. 

Even with state-owned enterprises, the central planners didn't try to centrally manage everything, they just didn't have the data (a knowledge problem). Generally these enterprises were operated by lower-level agencies, at more local levels, the central planners (Gosplan) set general targets and quotas and then the local agencies were meant to do the detailed planning within this. They were also meant to cover their own expenses out of their own revenue from sales, or temporarily by loans from the state bank (Bergson, page 212). The enterprises could also retain some of their profits, as an incentive for their employees. 

So the state didn't directly own all the economic activity in Russia. Bergson estimated that in 1937, the government budget was only 36% of total national income (Bergson followed the standard national accounts practice of counting household-production-for-own-use as part of national income).  Thus, the Soviet state used taxes, on the earnings of collectives, cooperatives, state-owned enterprises, and presumably those professionals with private practices. The main tax was actually a turnover tax (basically on sales). Applying an income tax as well did aim to capture those who earned income not caught by the turnover tax (e.g. small scale production), and it allowed a differentiation of being able to tax more those more able to afford it - there was a basic exemption up to X rubbles, exemptions for people living in remote areas, filing patents, exemptions for kids (and a bachelor levy). Replacing the income tax fully with a turnover tax would haven't allowed this flexibility around ability to pay, and thus reduced the amounts that the government raised - under Khruschev there was an attempt to eliminate the income tax but it 

On the consumer side, firstly consumer demand was mainly regulated not by prices but by shortages. No matter what the demand, consumers can't consume something that doesn't exist. The Soviet economy was known for shortages and sporadic availability. That said, if something was available, it was valuable to have the rubbles to buy it, so consumers found some value in rubbles and rubbles had value in the black market (though it was dangerous to be visibly rich). From the consumer point of view, money was a way of storing value. The state actually ran a lottery, and also issued state bonds that paid interest (sporadically, and with some ancellations), indicating a general demand by the population for money.

So the answer to your question is a combination of the state needing actual real resources to fund its activities, and the Soviet economy being rather more complex, due to both incentive and knowledge problems, than the idea of central planning suggests.

Sources

Holzman, F, 1958, Income Taxation in the Soviet Union: A Comparative View, National Tax Journal, http://www.jstor.org/stable/41790717

Hessler, J. (1998). A Postwar Perestroika? Toward a History of Private Enterprise in the USSR. Slavic Review, 57(3), 516-542. doi:10.2307/2500710, https://www.jstor.org/stable/2500710?read-now=1&seq=27#page_scan_tab_contents  

Bergson, A. (1950). Soviet National Income and Product in 1937. The Quarterly Journal of Economics, 64(2), 208-241. Retrieved February 7, 2020, from www.jstor.org/stable/1882693

Campbell, Robert. 1985. The conversion of national income data of the U.S.S.R. to concepts of the system of national accounts in dollars and estimation of growth rate (English). Staff working paper ; no. SWP 777. Washington, DC : The World Bank. http://documents.worldbank.org/curated/en/447651468743951934/The-conversion-of-national-income-data-of-the-U-S-S-R-to-concepts-of-the-system-of-national-accounts-in-dollars-and-estimation-of-growth-rate