A crash with core dump (very rare) : We will need the core dumps to be collected.
A crash due to sporadic issues, even things like a faulty hardware of disc full
Crash due to incorrect parameters or change in genesis values

Some thoughts on the time frame is given, please do share thoughts:

If there is a coredump, it needs to be collected. So the first step is to identity the presence of a coredump ( 15 minutes), at this point decision to spawn a new, but identical environment must be made. (OS, hardware, software versions everything must remain the same). A new environment can be then booted up in 2 hours (restore VM backup, start a fresh chain). All the accounts like escrow accounts or funding accounts must remain the same and the URL (domain) of the testnet must remain the same so that all connected testing environments works seamlessly. (Some of them like faucet will need a restart.) Over all time taken 3 hours

Since there is no core dump, this scenario requires recreating the TESTNET. Inspection 15 minutes, bringing up the testnet 30 minutes. Total time 45 minutes

The RC here is something like we had this week (genesis changes). This should warrant and immediate recreation of the environment. Inspection 1 minute and bringing up the environment 30 minutes.

Scenarios: 1. A crash with core dump (very rare) : We will need the core dumps to be collected. 2. A crash due to sporadic issues, even things like a faulty hardware of disc full 3. Crash due to incorrect parameters or change in genesis values Some thoughts on the time frame is given, please do share thoughts: 1. If there is a coredump, it needs to be collected. So the first step is to identity the presence of a coredump ( 15 minutes), at this point decision to spawn a new, but identical environment must be made. (OS, hardware, software versions everything must remain the same). A new environment can be then booted up in 2 hours (restore VM backup, start a fresh chain). All the accounts like escrow accounts or funding accounts must remain the same and the URL (domain) of the testnet must remain the same so that all connected testing environments works seamlessly. (Some of them like faucet will need a restart.) Over all time taken 3 hours 2. Since there is no core dump, this scenario requires recreating the TESTNET. Inspection 15 minutes, bringing up the testnet 30 minutes. Total time 45 minutes 3. The RC here is something like we had this week (genesis changes). This should warrant and immediate recreation of the environment. Inspection 1 minute and bringing up the environment 30 minutes.

rilesdun commented

2022-04-05 07:39:18 +00:00

(Migrated from gitlab.com)

This ticket is quite old - but I will give my 2 cents here; with current setup scripts the longest part for a redeployment is a fresh vm clone or restoration from backup.

Majority of crashes I have seen produce a core dump, but restoration time should never exceed a few hours under any circumstance.

This ticket is quite old - but I will give my 2 cents here; with current setup scripts the longest part for a redeployment is a fresh vm clone or restoration from backup. Majority of crashes I have seen produce a core dump, but restoration time should never exceed a few hours under any circumstance.

rilesdun commented

2022-04-05 07:39:47 +00:00

(Migrated from gitlab.com)

This ticket can be marked as closed - Along with the other related issues here; https://gitlab.com/PBSA/peerplays/-/issues/200 https://gitlab.com/PBSA/peerplays/-/issues/199

Please review @hbelakon @bobinson

This ticket can be marked as closed - Along with the other related issues here; https://gitlab.com/PBSA/peerplays/-/issues/200 https://gitlab.com/PBSA/peerplays/-/issues/199 Please review @hbelakon @bobinson

hbelakon (Migrated from gitlab.com) closed this issue

2022-05-04 17:25:25 +00:00