globaldiver
Contributor
nereas:
If you read a little closer, you see that it was a quote from a previous post from Ekewaka....
My reply was below that quote.
New glasses maybe?
Anders
horizontal ascents
- Thread starter nereas
- Start date
Please register or login
Welcome to ScubaBoard, the world's largest scuba diving community. Registration is not required to read the forums, but we encourage you to join. Joining has its benefits and enables you to participate in the discussions.
Benefits of registering include
- Ability to post and comment on topics and discussions.
- A Free photo gallery to share your dive photos with the world.
- You can make this box go away
If you read a little closer, you see that it was a quote from a previous post from Ekewaka....
My reply was below that quote.
New glasses maybe?
Anders
OP
nereas
Contributor
- Messages
- 2,735
- Reaction score
- 8
- # of dives
- 500 - 999
lamont:
This is what a tech instructor would warn everyone about as well, on the topic of argon.
Where is Chickdiver when we actually need her?
Soggy
Contributor
nereas:
What would you like to know from her?
JeffG
Contributor
Why CCR's are death boxes.Soggy:
cmalinowski
Contributor
ask an actuaryJeffG:
lamont
Contributor
lamont:
I dunno if I'm stretching here, but I see a really similar kind of mentality at work and in diving...
If everything goes right and its a sunny day you have nothing to worry about. You won't breathe any argon. The filesystem won't be full. Nobody will ever rewrite that startup/shutdown script so that it can hang. You don't need to code timeouts, you don't have to check exit codes.
I just took a 50 line bash script and turned it into a 450 line perl script mostly through adding timeouts and return code checking with a bunch of logging of its progress. I still haven't added any code to add a simple persisted on-disk state machine to the script so that it can do recovery when it gets interrupted and there's a bunch more sanity checking that could be done still. It'll probably be about 1,000 lines when it's done but it'll be considerably more bullet proof and it'll ***** like crazy when anything goes wrong.
And a lot of it may look excessively paranoid. Right now the application shutdown scripts just touch a file (which the app should read if it is operating) then sleep for a few seconds and kill -9 the process if its hung so badly it can't shutdown. That process should never go badly so there's no point wrapping my call to that shutdown script with a timeout -- but I can't guarantee in the future that nobody pushes out a version of the shutdown script with a hang in it. I want to be as future-proof as possible against anything bad happening.
The other approach to writing scripts like this is to code it like it'll always be a sunny day and then fix little issues in it whenever they come up. A couple of days ago they removed an infinite loop they ran into while waiting for connections to bleed off that never did which didn't just fail a single server but hung the entire platform. There's other infinite loops with no timeouts still in the codebase that I haven't removed that can still get triggered, though (in fact they only removed a condition that would cause a single server to hang, but didn't remove the infinite loop which also hung the whole platform when any single server hangs -- its just waiting for a single server to figure out another way to hang again).
I see strong parallels to approaches to technical diving. Some people tend to assume it'll be a sunny day all the time... I like my **** to work flawlessly no matter how cloudy it gets...
I am completely at a loss as to why you are looking for me?nereas:
OP
nereas
Contributor
- Messages
- 2,735
- Reaction score
- 8
- # of dives
- 500 - 999
chickdiver:
We need a really good tech instructor to answer some of the questions and speculations about argon.
JeffG
Contributor
We do??? You mean, you do.nereas:
OP
nereas
Contributor
- Messages
- 2,735
- Reaction score
- 8
- # of dives
- 500 - 999
I said, a really good tech instructor.
Similar threads
