We've had problems with boxes becoming unresponsive and cycling (no logs, hanging SSH sessions.) On inspection, the running program was taking up a significant amount of memory on the micro instance, which we think might be the cause:
This PR bumps the instance size for the manager from micro to small. Since deploying to CODE, I've been unable to replicate the problem, and there's more free memory on the box:
How to test
Tested by deploying to CODE. Since then, we haven't had a box cycle – I've given the test box a rinse with apachebench and I can't make it fall over, which is a promising sign (you'll need to replace the cookie with something current):
ab -n 100 -c 10 -H 'accept: */*' -H 'accept-language: en-GB,en-US;q=0.9,en;q=0.8' -H 'accept-language: en-GB,en-US;q=0.9,en;q=0.8' -H 'cookie: _ga=GA1.3.545298348.1685975974; _gid=GA1.3.1852050354.1688377328; consentUUID=3cda161a-1282-4683-b81c-3fe8e68de5f5_21; consentDate=2023-07-05T09:26:05.167Z; features.override=dev-mode%3Dtrue; gutoolsAuth-assym=<add your panda cookie here>' -H 'sec-fetch-site: same-origin' -H 'user-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36' 'https://manager.typerighter.code.dev-gutools.co.uk/rules'
What does this change?
We've had problems with boxes becoming unresponsive and cycling (no logs, hanging SSH sessions.) On inspection, the running program was taking up a significant amount of memory on the
microinstance, which we think might be the cause:This PR bumps the instance size for the manager from
microtosmall. Since deploying to CODE, I've been unable to replicate the problem, and there's more free memory on the box:How to test
Tested by deploying to CODE. Since then, we haven't had a box cycle – I've given the test box a rinse with apachebench and I can't make it fall over, which is a promising sign (you'll need to replace the cookie with something current):