0

Would the SQL Server instance go offline if it couldn't write to any of the Windows Event logs?

Found these two errors in the Windows Systems Event Log:

The system failed to flush data to the transaction log. Corruption may occur.

initerrlog: Could not open error log file 'E:\Program Files\Microsoft SQL Server\MSSQL10_50.MSSQLSERVER\MSSQL\Log\ERRORLOG'. Operating system error = 5#Access is denied.#.

From what I can gather, these issues are related to our SQL Server instance going offline completely.

I've verified the security of the SQL Server Agent and Database Engine service accounts have the required access to the aforementioned log path. There is nothing in the SQL Server Error logs or the Windows Logs that state things were changed by someone or that that they automagically lost their access some how.

Let me know if you need more info and thank you in advance for any pointers you can provide.

AKDiscer
  • 123
  • 2
  • 11
  • 1
    Is there enough disk space available on the E drive? How often do you get these errors? Please also paste the full event details, at least including event source and event id. – Lucky Luke Aug 29 '16 at 16:30
  • @LuckyLuke the error message was from the Windows Application Log with the Event ID of 17058. This is the first time we've received this error and there was never a space issue. The error message listed was all that was in the event details. – AKDiscer Aug 30 '16 at 19:06

1 Answers1

0

I'm not going to be able to completely answer your question, but hopefully I can get you moving closer to the path you need to be on. I'm pretty sure that those are not Windows Event Logs but the event logs within MSSQL itself. Your database is in full recovery mode it sounds like, and most likely you have ran out of space or the permissions have become borked on the transaction log files according to the error you posted. If you're out of space you will have to switch over to simple backup mode and take a backup so the transaction log files can be flushed and space can be recovered. If you happen to just be able to magically expand space on the drive that would be great too. Once you get past this immediate issue you still need to have a maintence task that backs up your DBs so the logs can be cleared.

JBaldridge
  • 484
  • 3
  • 10
  • There is nothing in the SQL Error Logs because they couldn't be written to. There is a gap during the time that the instance went offline, there is only information in the Windows Event Logs, the Application log is more telling that the System Log, there wasn't much of any type of bread crumb to follow. Also, there was no space issue on this drive, there was tons of space available. Further, the MDF and LDF containers had plenty of free space and there is no limit to their growth. – AKDiscer Aug 30 '16 at 19:08
  • I have a real suspicion that this drive went offline in the SAN somehow, but I can't prove it one way or another. I'd like to think that there are logs in the SAN that would state this, but I could be wrong. Still waiting on a reply from my SAN/Storage admin. – AKDiscer Aug 30 '16 at 19:09
  • Ok, I apologize but I read that wrong. It's been a while since I have dealt with a SAN so I am not too sure what logs to suggest to look at to verify connectivity. Did you try looking in the system logs about that time? There could also be something buried in the :Applications and Services Logs. Here's an MSKB that talks about system logs to look for: [2091098](https://support.microsoft.com/en-us/kb/2091098). – JBaldridge Aug 30 '16 at 19:38