fix(azure event hub): wrong calc of the remaining items

[yodobrin]

as the scenario covered in this line is for the case in which the partition reached maxInt and started from zero. the calc is done wrongly.

Signed-off-by: Yoav Dobrin 37622785+yodobrin@users.noreply.github.com

the calc of the remaining items to process is wrong

Checklist

• Commits are signed with Developer Certificate of Origin (DCO - learn more)

[JorTurFer]

/run-e2e event_hub*
Update: You can check the progress here

[JorTurFer]

Thanks for the contribution! ❤️
LGTM, but I'm not an expert in EH, @v-shenoy , do you know if this change is correct?

[yodobrin]

The most common scenario would be that the partition.lastseq would be larger than the checkpoint.seq – this is addressed in previous lines of code.
Since the partition is a cyclic counter, once it reaches maxInt it will zero, in that scenario the checkpoint.seq will be larger.
Since we try to calc the number of records we need to process

Lets say max is 100
Partition.lastseq is 3
Checkpoint.seq is 6

How many records do we need to process?
In the current calc we have 100-3+6=103
In the suggested we have 100-6+3=97

Since the number of records from 6 to 100 is 94, and the number of newly created records is 3, the number of records we need to handle is 97, not 103.

[JorTurFer]

The most common scenario would be that the partition.lastseq would be larger than the checkpoint.seq – this is addressed in previous lines of code. Since the partition is a cyclic counter, once it reaches maxInt it will zero, in that scenario the checkpoint.seq will be larger. Since we try to calc the number of records we need to process

Lets say max is 100 Partition.lastseq is 3 Checkpoint.seq is 6

How many records do we need to process? In the current calc we have 100-3+6=103 In the suggested we have 100-6+3=97

Since the number of records from 6 to 100 is 94, and the number of newly created records is 3, the number of records we need to handle is 97, not 103.

I have been checking your numbers and makes sense in my mind (I'm not really good at mathematics imaging shapes TBH), let's way till other folk confirms this and we can merge it.
Thanks for the sample you have provided, it's really useful

[v-shenoy]

Thanks for the contribution! ❤️
LGTM, but I'm not an expert in EH, @v-shenoy , do you know if this change is correct?

I will spend some time on this later today, and get back.

[yodobrin]

Hi @v-shenoy and @JorTurFer - please update with your decsion

[JorTurFer]

/run-e2e event_hub*
Update: You can check the progress here

[JorTurFer]

/run-e2e event_hub*
Update: You can check the progress here

[JorTurFer]

LGTM!
Let's wait some time for giving the option to @v-shenoy to check this if he has time, and if not I'll merge it in a few days

[v-shenoy]

LGTM! Let's wait some time for gibing the option to @v-shenoy to check this if he has time, and if not I'll merge it in a few days

I am not completely sure about this either, but this seems right.