Attention is currently required from: neels, laforge, pespin. iedemam has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-bsc/+/27081 )
Change subject: WIP: New stats for lchan life duration. ......................................................................
Patch Set 7:
(4 comments)
Patchset:
PS7: Hi all,
More updates. I fixed type usage when pulling rate_crt and stat_items as well.
This (unfortunately?) led me to think a rate_crt is indeed the better type for storing total ms. Simply because it is a much larger type and will only ever increase. We will never have fewer active ms from one period to another.
Thoughts? How is wrapping handled in a stat_item? If I'm calculating correctly, we will wrap the ms value there after approximately 600 Erlangs. No problem if it's handled gracefully...I just want to point it out.
Thanks, -Michael
File src/osmo-bsc/bts_trx_vty.c:
https://gerrit.osmocom.org/c/osmo-bsc/+/27081/comment/9a580c6c_f6a17e38 PS4, Line 575: vty_out(vty, " Activated %llu ms ago%s", duration_ms, VTY_NEWLINE);
Feel free to print it as ms if value < 1000. Or print them as "X. […]
Now I show "<1s" if sub 1000ms and "Xs" otherwise.
File src/osmo-bsc/bts_vty.c:
https://gerrit.osmocom.org/c/osmo-bsc/+/27081/comment/0eff2fad_081ffdcf PS6, Line 3831: vty_out(vty, " TCH %"PRIu64"", activations_tch);
I ment the left side tabulation ;) It's 2 spaces per indentation leel afaict
Sorry, gotcha. I was right-aligning TCH and SDCCH. It should be correct now.
File src/osmo-bsc/lchan_fsm.c:
https://gerrit.osmocom.org/c/osmo-bsc/+/27081/comment/d3d793d6_b741daad PS5, Line 1815: case GSM_LCHAN_TCH_F:
we put "case" line in the same column as "switch" in osmocom syntax style.
aligned