Added silk_request_garbage_collect command for out-of-band garbage collection.

[nschlemm]

To avoid deadlock issues as mentioned e.g. in #265 #294 #371
Based on #265 (comment)

[codecov]

Codecov Report

Merging #541 (a785e53) into master (be920b2) will increase coverage by 0.11%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #541      +/-   ##
==========================================
+ Coverage   84.60%   84.72%   +0.11%     
==========================================
  Files          51       52       +1     
  Lines        2059     2075      +16     
==========================================
+ Hits         1742     1758      +16     
  Misses        317      317              

Impacted Files	Coverage Δ
...anagement/commands/silk_request_garbage_collect.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update be920b2...a785e53. Read the comment docs.

[hramezani]

Hi @nschlemm, Thanks for mentioning me in this patch and asking for a review.
Unfortunately, I am not involved in this project.
So, you may need to ask project leads for a review.

[nschlemm]

@hramezani sorry, then github must definitely step up their game for suggesting reviewers ;)

and I can't see an option to "release" you from this mention...

@auvipy as release author of 4.2.0 would you be willing to review this? thx!

[albertyw]

Even when triggered manually through this command, it's still possible for deadlocks to happen unless the database is receiving no other writes during the period. This isn't going to guarantee fixing deadlocks. There should be a full solution made to fix deadlocks, though the ability to manually trigger cleanup perhaps during low-traffic hours is still useful.

[albertyw]

There are some rebase conflicts. @nschlemm Please rebase and this can be merged.