1

Topic: Prompt about data migration between servers. Import is eliminated, unless in a file.

There are some servers, more than three, dataful. With similar on a format, but the data different in the content.
It is necessary, with certain periodicity, to select a certain data set, from servers of sources and to add result of request in the table on absolutely other server, the receiver.
As it is necessary to eliminate as much as possible from this operation of people, in sense on a maximum to make means of jobs of the agent.
While I see only results of requests to add in XML, to transfer XML files to the server the receiver and on it from XML to transfer the data to basis the receiver.
Can people already solved similar tasks?
The full access to servers to sources is not present, only the receiver. But to ask/force administrators to make  and procedure it it is possible, as scripts of requests I will give.

2

Re: Prompt about data migration between servers. Import is eliminated, unless in a file.

On the bad communication lines
https://docs.microsoft.com/ru-ru/sql/re … rogramming
https://docs.microsoft.com/ru-ru/sql/re … ent-studio
https://docs.microsoft.com/ru-ru/sql/re … hrough-ftp

3

Re: Prompt about data migration between servers. Import is eliminated, unless in a file.

sfyk;
Replication? SSIS-> Transfer Database Task and Transfer SQL Server Objects Task? bcp? SSIS-packages?

4

Re: Prompt about data migration between servers. Import is eliminated, unless in a file.

BOBAHbI4;
It is required not less than three flash cards and the reliable car.

5

Re: Prompt about data migration between servers. Import is eliminated, unless in a file.

BOBAHbI4, and what price of a question? How many it is necessary to transfer the data and with what periodicity?
I, for example, collect thus about 2-3 Gb of the data from each of 36 servers each 4 hours.
On the server a source - bcp with result of request in native a format, then - I archive zip th (they are pressed time in 8), I transport with the help syncthing, on the server the receiver - I unclench, and then - insert into... from openrowset (bulk... formatfile...) with (tablock)
I load in the separate table corresponding to section, and then I do switch partition.
(2016SP1 supports the partitioned tables in standard versions).
Basically, all normally and very quickly works even with Chukotka and Magadan, with which channel 256 on the companion.
If versions - more ancient, and the "instant" update of section in the general table on the server-receiver - nevertheless is necessary, I would advise to get set of the tables corresponding to servers to sources, to unite them with the help union all, given to load in  similar tables, and then to replace out-of-date tables which are included in view union all with the help switch table.