Block production summary in 100 node #1876
Conversation
Codecov Report
@@ Coverage Diff @@
## staging #1876 +/- ##
==========================================
Coverage ? 86.76%
==========================================
Files ? 176
Lines ? 34153
Branches ? 0
==========================================
Hits ? 29634
Misses ? 4519
Partials ? 0Continue to review full report at Codecov.
|
|
Test result so far: block production time 10s, fork after about 4 hours block production time 100s: not fork after 12 hours. cpu 100% though block production time 55s: block production time 30s: block production time 40s: |
|
Maybe we are doomed. Which one do you refer to when you talk about block production time? Is it |
|
All three Paramus. For 10s, it means these three * 10
…On Sun, Dec 22, 2019 at 11:33 AM Bowen Wang ***@***.***> wrote:
Maybe we are doomed. Which one do you refer to when you talk about block
production time? Is it min_block_production_delay,
max_block_production_delay, or max_block_wait_delay?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#1876>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ADFFFCHTQYVWZ47ZVMHGZY3QZ66GXANCNFSM4J4DPPAA>
.
|
|
So |
|
Yes
…On Sun, Dec 22, 2019 at 11:36 AM Bowen Wang ***@***.***> wrote:
So min_block_production_delay is 10s, max_block_production_delay is 20s
and max_block_wait_delay is 60s?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#1876>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ADFFFCGT6MGFNHP5ACJUTZTQZ66SZANCNFSM4J4DPPAA>
.
|
|
That's unfortunate. Is memory usage normal? |
|
Memory is under 1G in the longest test of 12 hours, not sure of longer
test. Once current 40s test finish I’ll pay attention to memory tendency in
log
…On Sun, Dec 22, 2019 at 11:37 AM Bowen Wang ***@***.***> wrote:
That's unfortunate. Is memory usage normal?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#1876>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ADFFFCGHT5UQR7SFQUXUC6TQZ66X5ANCNFSM4J4DPPAA>
.
|
|
Memory isn't very optimistic, gradually increase to 1.3G and looks still
increasing, I attached a copy of log from two nodes. All node looks
similar, but node17 is special coz all of validator /status show only
node17 to be validator now.
…On Sun, Dec 22, 2019 at 11:41 AM Bo Yao ***@***.***> wrote:
Memory is under 1G in the longest test of 12 hours, not sure of longer
test. Once current 40s test finish I’ll pay attention to memory tendency in
log
On Sun, Dec 22, 2019 at 11:37 AM Bowen Wang ***@***.***>
wrote:
> That's unfortunate. Is memory usage normal?
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <#1876>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/ADFFFCGHT5UQR7SFQUXUC6TQZ66X5ANCNFSM4J4DPPAA>
> .
>
|
|
yeah there seems to be some new memory leak issue. Will investigate separately. |
|
@bowenwang1996 @mfornet |
* docs: Improve README for one hundred nodes * test: Enable diagnostic in one hundred nodes test * doc: Add more instructions for the 100 nodes test
ailisp
requested review from
evgenykuzyakov and
MaksymZavershynskyi
as code owners
| node_blocks = [0] * 100 | ||
| block_by = {} | ||
| for i in range(100): | ||
| with open(f'pytest-node-{i}.txt') as f: |
There was a problem hiding this comment.
can we clean up this a bit? It looks outdated.
There was a problem hiding this comment.
Sure. The problem is I don't exactly remember which was outdated 😂 but I guess you and Marcelo has already keep one or two useful, up to date scripts, So I can simply delete all analyze scripts here?
There was a problem hiding this comment.
That's also fine I guess
There was a problem hiding this comment.
Deleted. If you feel some of your analyze script help, please add to this branch
Output is roughly as above. Each line in each txt above is in format:
block_hash block_height approvals is_chunk_included_all_in_current_height chunk_included_height