Circuit breaker config not working properly - molecule

I am setting up new molecular project and trying to config circuit breaker in my project under molecular.config.js and set windowOpen to 6 sec. But when i do any operation and throw an error. Circuit doesn't break down. I am not able to find any solution for this.
Need Help

Here you can see the parameters: https://moleculer.services/docs/0.13/fault-tolerance.html#Settings
The default values don't trip the CB after the first error. E.g. the number of requests should reach the minRequestCount. If you want to trip the CB after the first error, set threshold: 1 & minRequestCount: 0

Related

PedSelectOutPut routing pedestrians inconsistently

I am simulating the passenger changeover process in metros using the Anylogic Pedestrian Library.
When passengers enter the vehicle, a seat is assigned to them from the seats available near the door (within a given distance) they entered the vehicle through, using a function called lookForSeat. If there is no more free seat available, their boolean parameter wantToSit is set to false and they will stay standing.
The parameter wantToSit is predefined for the Passenger Agent, with default value randomtrue(0.8). But even if I set it to default value = 1, I get the same error.
Then, passengers are separated using a PedSelectOutput block:
Condition 1: if ped.WantToSit = true --> they are sent to their
assigned seat coordinates (PointNode 'seatPoint', null by default)
Condition 2: true (thus, ped.WantToSit = false) --> they stay in the
standing area in the vehicle, no assigned seatPoint necessary in this case.
Now, it works pretty well, but when the majority of the seats is already occupied, suddenly the PedSelectOutput block directs a passenger with ped.wantToSit to its seating point, which gives null and I get the NullPointerException error.
Attached you find the function, the settings of PedSelectOutput and the log from the command.
As it can be seen, the PedSelectOutput sends the passenger through exit 1 (which gives the error due to calling the coordinates of a "null"), despite ped.wantToSit = false.
Any ideas, what is going wrong? For me it really looks like the function is working properly - I have been changing it the whole day until I realized that something in the PedSelectOutput block goes wrong.
Thank you in advance!
Pic 1: pedSelectOutput block and the command with the log
Pic 2: the function lookForSeat assigning the seats from the seat Collection
The problem here is a subtle one, which has caused me many hours of debugging as well. What you need to understand is that the on exit code is only executed once the agent already has a path to which it is going to exit. i.e. the selectOutput and subsequent blocks are already evaluated and only once it is determined that the agent can move to the next block then the on exit code is called. But the agent will continue on its chosen path that has been determined before the on exit code has been executed.
See the small example below:
I have a pedestrian with a variable that is true by default and a select output that checks this value
If I ran the model all pedestrians exit at the top option, as expected
If I change the variable to false on the On Exit code I might expect that all pedestrians will now exit at the second option
But they don't there is no change....
If I add the code to the on enter code then it does..

Juliaopt JuMP CbcSolver how to print progress

I coding in julia using JuMP of Juliopt.
I would like to know how to print the solve progress using CbcSolver?
I already used the log level parameter:
m = Model(solver = CbcSolver(log=1))
it works and you can choose between 0 to 3, higher more details.
but my problem is that the log level only prints after the solver finish, i already put the the time limit parameter "sec" but when it is set for higher values never stops and i don't know why!!
m = Model(solver = CbcSolver(sec=90,log=1))
I was going to try callbacks but Cbc does not support callback.
I am positive that there is a way to print because when i use the opensolver in excel a set Cbc to solve it shows the progress.I just dont know how to do this!!

AUTOSAR configuaration - DCM module

I am stuck at a point where I am configuring the DCM module and the current parameter I am trying to configure DcmTimStrP2AdjustServer,
The requirement is P2CAN_SERVER_MAX = 25ms; P2STARCAN_SERVER_MAX = 5000ms;
Is DcmDspSessionP2ServerMax the same as P2CAN_SERVER_MAX? and if it is the same
What is the need for DcmTimStrP2AdjustServer and how do I find the best value for DcmTimStrP2AdjustServer.(The values all should be a multiple of DcmTaskTime which I find to be logical).
DcmTaskTime = 5ms;
I am following Autosar 4.0.3, using ETAS tool for configuring the parameters.
To fulfill your requirement, you need to configure respectively
DcmDspSessionP2ServerMax & DcmDspSessionP2StarServerMax for each session control in the DcmDspSessionRows at Dcm/DcmConfigSet/DcmDsp/DcmDspSession/.
i.e.
DcmDspSessionP2ServerMax 25
DcmDspSessionP2StarServerMax 5000
There is no DcmTimStrP2AdjustServer, but I guess you're referring to DcmTimStrP2ServerAdjust instead. DcmTimStrP2ServerAdjust & DcmTimStrP2StarServerAdjust should be configured to a multiple of your DcmTaskTime (5ms in your case, so i.e. 5ms, 10ms, 15, ms, ... is applicable) and are used to safeguard that the response is available on the bus before triggering the P2 or P2* timeouts. In your case you may want to set these values to the same values as in the DcmDspSessionRows if there is no other specification given, because the chosen timeout values there are already multiples of your DcmTaskTime:
DcmTimStrP2ServerAdjust 25
DcmTimStrP2StarServerAdjust 5000
The adjust value is an internal value, in order to adjust the delay between the Dcm Transmit Request and the message being actually on the Bus.
The definition of P2ServerMax and P2*ServerMax and their corresponding Adjust values is the same:
This parameter is used to guarantee that the diagnostic response is available on the bus before reaching P2 by adjusting the current DcmDspSessionP2ServerMax. This parameter mainly represents the software architecture dependent communication delay between the time the transmission is initiated by DCM and the time when the message is actually transmitted to the bus

cocoa-applescript: running handler or command every few seconds

In normal applescript, the script is executed down the page, and so any code in loops for every 5 seconds will only run while the loop is running - there is no way to have a single function run every few second regardless of what the script is currently doing or where it is in the script (that I know of). In cocoa-applescript, however, is there a way to run a handler every 5 seconds, at all times, no matter what it is currently doing? Here is what it should be doing in my cocoa-applescript app:
on checkInternetStrength()
do shell script "/System/Library/PrivateFrameworks/Apple80211.framework/Versions/Current/Resources/airport -I | grep 'agrCtlRSSI:'" -- this being the script which returns the line containing the signal strength
set SignalStrength to result
set RSSIcount to (count of characters in SignalStrength)
set SignalStrength to ((characters 18 thru RSSIcount of SignalStrength) as string) as integer -- this to turn SignalStrength into just the number and not the whole output line
set SignalStrength to (100 + SignalStrength) as integer
set SignalBar's setIntValue_(SignalStrength) -- SignalBar being the Level Indicator described below
end checkInternetStrength
Summed up, it runs the airport command to check internet connection, turns this into a number from 1 to 100 and uses this on an NSLevelIndicator (100 maximum) to show current signal strength graphically. Now, there is no point having this run once or when you hit a button - that is an option, but it would be nice if it updated itself every, say, 5 seconds with the realtime value. So is there any way to have a process which runs every 5 seconds to do this, while still enabling full functionality of the rest of the script and interface - i.e. as a background process? Comment if you need more extracts from the script.
Example
In Unity-C# scripting, the 'void Update() {code}' will run the code within it every frame while doing everything else simultaneously, so a cocoa-applescript version of this might be an answer, if anyone knows.
I Dont believe this is possible but what I had a similar problem before, what i do, I have an external applescript applicaion that is hidden the repeats the commands, the only problem is, it wont send it back to the app, you'll have to make the external applescript app do it, like
display notification, etc..., in the applescript apps "Info.plist" you can add this:
<key>LSUIElement</key>
<string>1</string>
To make the app run invisibly, but sorry i dont think you can run a handler in the app its self

Is it possible to set dynamic download delay in scrapy?

I know that a constant delay can be set in
settings.py
DOWNLOAD_DELAY = 2
however, if I set the delay to 2s it is not efficient enough. If I set the DOWNLOAD_DELAY = 0.
The crawler is able to crawl about 10 pages. after that, the target page will return something like " you are requesting too frequently ".
What I want to do is the keep the download_delay to 0. once the "requesting too frequently" msg is found in the html. it change the delay to 2s. After a while it switch back to zero.
is there any module can do this? or any other better idea to handle such case?
Update:
I found that is a extension call AutoThrottle
but is it able to customize some logic like this??
if (requesting too frequently) is found
increase the DOWNLOAD_DELAY
If right after you get anti-spider page, then in 2 seconds you can get data page, then what you are asking probably requires writing a downloader middleware
that checks for anti-spider page, reset all scheduled requests to a renew-queue, start a looping call when spider is idle to get request from the renew-queue, (the looping interval is your hack for a new download delay), and try to decide when the download delay is not necessary again (requires some tests), then stop the looping and reschedule all the requests in renew-queue to scrapy scheduler. You will need to use redis queue in case of distributed crawl.
With download delay set to 0, in my experience throughput can go easily above 1000 items/min. If anti-spider page pops up after 10 responses, then it is not worth the effort.
Instead maybe you can try to find out how fast does your target server allow, may be 1.5s, 1s, 0.7s, 0.5s etc. Then maybe redesign your product takes into consideration the throughput your crawler can achieve.
You can use Auto Throttle extension now. It is turned off by default. You can add these parameters in your project's settings.py file to enable it.
AUTOTHROTTLE_ENABLED = True
# The initial download delay
AUTOTHROTTLE_START_DELAY = 5
# The maximum download delay to be set in case of high latencies
AUTOTHROTTLE_MAX_DELAY = 300
# The average number of requests Scrapy should be sending in parallel to
# each remote server
AUTOTHROTTLE_TARGET_CONCURRENCY = 1.0
# Enable showing throttling stats for every response received:
AUTOTHROTTLE_DEBUG = True
Yes, You can use the time module to set the dynamic delay.
import time
for i in range(10):
*** Operations 1****
time.sleep( i )
*** Operations 2****
Now you can see the delay between Operations 1 and Operations 2.
Note:
the variable 'i' is in the form of seconds.