Timeline for How should I handle logger failures?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 13, 2014 at 5:48 | comment | added | Aaronaught | What if you end up with timeouts sending to syslog or I/O errors writing to a file? You could still be making the problem worse, if the failures are due to a congested network or running out of disk space. This isn't exactly a holistic solution; you need to consider the possibility that there may not be any safe way of logging the errors. It's not that dangerous to log to your own logger as long as you incorporate cycle detection, exponential back-off, etc. | |
| Dec 12, 2014 at 20:27 | comment | added | Zairja | I think we will try to implement a fallback as you suggest. Jon Raynor's suggestion to stop the application (in a critical logging situation) is also one we may pursue that we hadn't considered. | |
| Dec 12, 2014 at 20:25 | vote | accept | Zairja | ||
| Dec 12, 2014 at 17:45 | history | edited | Arseni Mourzenko | CC BY-SA 3.0 | added 1107 characters in body |
| Dec 12, 2014 at 17:41 | comment | added | mgw854 | I ran into this same problem with my exception logger that went to e-mail. If it failed to connect to a server, it got into a terrible infinite loop. So instead, I put a check in place to divert to the Event Log and prevent new e-mails from being sent out until a new connection could be made. | |
| Dec 12, 2014 at 17:38 | history | answered | Arseni Mourzenko | CC BY-SA 3.0 |