feat: replace github.com/denisenkom/go-mssqldb with github.com/microsoft/go-mssqldb

[powersj]

fixes: #12745

[powersj]

@minwal, @AsafMah,

I need some help ensuring this PR does not regress you or our users of the azure data explorer output. This PR updates the go-mssqldb from github.com/denisenkom/go-mssqldb to github.com/microsoft/go-mssqldb.

Most of the changes are drop in except the following commit to azure data explorer output: 00cb07e

There are two changes I am concerned about:

1. The way we login. I did not see a drop in replacement for NewAuthorizerFromEnvironmentWithResource. The WithDefaultAzureCredential does state that it looks at "The order of auth will be via environment variables, managed identity and Azure CLI." Which I think aligns with the current behavior.
2. The create table and table mappings, kuso.NewStmt is deprecated and replaced with kql.New(). I was not sure if my updates are enough given I am missing the suppress warnings statement that was previously there.

Can either of you try this out or verify that I have made the correct changes?

@soccypowa,

Since you were also interested in the replacement of the go-msssqldb dependency, can you give this a shot to verify your kerberos support works?

Thanks!

[soccypowa]

@powersj

I made a build of your branch and took it for a spin. For some reason though I got only NTLM auth in my test environment, realised that I get the same behaviour with my own updated "ugly build" go-mssqldb v1.3.0, while my old one, made in the excact same way with go-mssqldb v0.17.1-0.20221011173445-537893aca440 just works. I tried to do a fast "read up" on wether MSFT changed some behaviour or not, I mean they have changed default settings in many of their other drivers so why not this one. Have not found any such indications, but if my working schedule allows me I will look into this, in any case out of curiosity. Nice work getting the change in place though!

Update:
Ok, seemingly it is isolated to one of my hosts, the others get Kerberos as expected. So in other words this looks great!

[srebhan]

Looks good to me. We should try to remove the go.mod replacement as soon as possible IMO...

[srebhan]

Is there any PR we can watch to remove this replacement ASAP?

[srebhan]

@powersj giving this back to you in case you want to wait for the upstream PR to be merged...

[powersj]

@srebhan Upstream PR is merged, thanks for catching the replace was still there. I've updated the go.mod to use the latest commit and removed the replace.

[telegraf-tiger]

Download PR build artifacts for linux_amd64.tar.gz, darwin_amd64.tar.gz, and windows_amd64.zip.
Downloads for additional architectures and packages are available below.

⚠️ This pull request increases the Telegraf binary size by 3.37 % for linux amd64 (new size: 193.7 MB, nightly size 187.4 MB)

📦 Click here to get additional PR build artifacts

Artifact URLs

DEB	RPM	TAR GZ	ZIP
amd64.deb	aarch64.rpm	darwin_amd64.tar.gz	windows_amd64.zip
arm64.deb	armel.rpm	darwin_arm64.tar.gz	windows_arm64.zip
armel.deb	armv6hl.rpm	freebsd_amd64.tar.gz	windows_i386.zip
armhf.deb	i386.rpm	freebsd_armv7.tar.gz
i386.deb	ppc64le.rpm	freebsd_i386.tar.gz
mips.deb	riscv64.rpm	linux_amd64.tar.gz
mipsel.deb	s390x.rpm	linux_arm64.tar.gz
ppc64el.deb	x86_64.rpm	linux_armel.tar.gz
riscv64.deb		linux_armhf.tar.gz
s390x.deb		linux_i386.tar.gz
		linux_mips.tar.gz
		linux_mipsel.tar.gz
		linux_ppc64le.tar.gz
		linux_riscv64.tar.gz
		linux_s390x.tar.gz