How to locate an obscure paper? A colleague asked me to locate the following paper on the web:
Kovalenko,  I.N.:  On  the  reconstruction  of  an  additive  type  of  distribution  based  upon  a sequence of  independent  trials.
Memoirs  of  the  All-Union Conference on  Probability Theory and Mathematical Statistics,  Erevan  1958.
After a few failed attempts, I asked myself, how would I go about locating this paper in the REAL (not virtual) world? What library stocks proceedings of the All-Union Conference... in Erevan, from 1958?
Then it occurred to me that of course, the paper had to be in Russian. Being a Russian speaker, it wasn't too difficult to recover the original title:
Коваленко И.Н., О восстановлении аддитивного типа распределения по последовательной серии независимых испытаний, Труды Всесоюзн. Совещания по теор. вероятн. и матем. стат., Ереван, 1958, Изд-во АН АрмССР, Ереван,1960.
So if any Russian-based readers out there have access to this archive, I would very much appreciate a PDF of this paper.
But I am more interested in the process. How does a researcher go about locating such things?
 A: I have two related stories, one is my MO question:
Does anyone have access to a copy of Yury G. Teterin's 1984 (Russian) preprint "Representation of numbers by spinor genera" 
The other was a 1930's article in Ukrainian, as I recall, that is not critical. I do not seem to have kept the email from that far back. Anyway, a friend of a friend of a friend was at a different university in Ukraine, and was willing to find the item and scan it in and email the pdf. 
One thing I did wrong with the latter item (but chronologically earlier) was to try to rope a bunch of people into contributing to translating the item, including A. Eremenko at Purdue. That did not work and was not appreciated. So, a translation or part of one is a bigger deal, by far, than an article scan. My friend Dmitry read me selected sentences by telephone but indicated it gave him a headache, some words being uncertain anyway as differing from Russian. As clumsy as I was, this was worth doing, nobody knew about this article.
So I guess my summary is, people who have some involvement with mathematics are pretty generous about scanning in articles for people they will never meet. However, it does not do to get greedy with the kindness of strangers. 
A: Locating the paper on MathSciNet is always a good idea.
First one should look at electronic mathematical libraries.
Quite often they have an electronic version of the book.
Another efficient method is to look for papers that contain references to the paper under consideration
and then ask the author of such a paper for a copy.
In our situation a search reveals just one such paper: Mattner, L., On Burdick’s symmetry problem., Math. Methods Stat. 8, No. 1, 109-118 (1999). ZBL1104.62302. Preprint available at CiteSeerX.
Unfortunately, the author explicitly states that he was not able to locate a copy of the paper that we are looking for.
If all else fails, one should do a search on WorldCat to locate libraries that have a paper copy.
In our case a search reveals quite a few libraries.
(When looking for a book whose original title is written in a non-Latin alphabet,
one should be extremely careful in formulating the request.  In our case the words
“trudy” and “statistike” seem to be most resistant to distortion by
transliteration systems used by different libraries, hence we enter
them into our search query.)
Numerous university libraries in the US, France, Germany, Denmark, Slovenia,
Czech Republic, Switzerland, Poland, and Australia seem to possess the book you are looking for.
Most probably your university has an interlibrary loan agreement with at least one of these universities.
Perhaps one of the libraries might be willing to scan the paper for you (maybe for some small fee),
you might have a friend in one of these universities,
or when traveling to a conference at one of these universities you can simply stop by the library.
Writing directly to Kovalenko (currently he seems to be affiliated with London Metropolitan
University)
and requesting a paper copy should also be an option. (Update: this is no longer possible, since Kovalenko passed away in October 2019.)
A: As for your last question: when I try to find "obscure" papers on the web, I mostly do a very broad search on zbMATH Open.
I tried searching for your paper searching for the author, only to find out he has over >100 papers (or perhaps multiple authors with the same name). Since you had the publishing year, finding the desired result was easy:

Kovalenko, I. N., Über die Wiederherstellung des additiven Verteilungstypus nach einer Folge von unabhängigen Beobachtungsserien, Tr. Vsesojuzn. Soveshch. Teor. Veroyatn. Mat. Statist., Erevan 1958, 148-159 (1960). ZBL0208.44802.

Sadly, in this case, zbMATH doesn't have a link to a free version of the paper.
But at this stage, I mostly have enough to go on to do a refined Google search. In this case, however, this also failed.
Emil's comment:

Ask in your library. They may not have it directly, but they can probably figure out which library has it and obtain a photocopy from them for some fee (or even the real thing as an interlibrary loan, if you are lucky).

would then be my next step, as well as asking authors who have referenced the paper.
A: A minor addition to the excellent answer by Dmitri Pavlov: the author's (other?) webpage appears to be hosted by the V.M. Glushkov Institute of Cybernetics of National Academy of Sciences of Ukraine, and his e-mail address is given there. (Update: contacting Kovalenko is no longer possible, since he passed away in October 2019.)
A: University of Yerevan has a copy (http://89.249.207.20/cgi-bin/koha/opac-detail.pl?biblionumber=156826&query_desc=kw%2Cwrdl%3A%20%D0%A1%D0%BE%D0%B2%D0%B5%D1%89%D0%B0%D0%BD%D0%B8%D1%8F%20%D0%BF%D0%BE%20%D1%82%D0%B5%D0%BE%D1%80.%20%D0%B2%D0%B5%D1%80%D0%BE%D1%8F%D1%82%D0%BD.%20%D0%B8%20%D0%BC%D0%B0%D1%82%D0%B5%D0%BC.%20%D1%81%D1%82%D0%B0%D1%82), and you can obtain a copy from them for a fee.
Contact their librarian directly (mirzoyan@ysu.am). I don't know at this stage in history whether English or Russian is the preferred international language of Armenian librarians, but you are clearly fluent in both.
A: One should not expect a 1958 paper to be on the web.  The author did not prepare it in machine-readable form, so he won't have it.  So it is probably available only if the publisher (or some third party) has gone to the expense of digitizing their back issues.
