How can we help you today? How can we help you today?

2.2 - Bogus "Log backup overdue" messages

2/18: Installed 2.2 - upgraded from 2.1.
2/20: Firewall reconfigured by ISP, which knocked out SQL Monitor.
2/22: raft of "Log backup overdue" messages emitted for all "Full" Recovery Mode DBs at the time that SQL Monitor was re-authenticated after the firewall hiccup. At the same time, SQL Monitor alerted me to an overdue FULL backup for a DB that had been configured to "Ignore Alert", as it only runs weekly and is a reference-only DB.

Tran Log backups for 2 of the featured DBs run every 15 minutes - hardly "Overdue". The others run hourly or 2-hourly. All were up to date at the time the alerts were raised.

OOPS?
PDinCA
0

Comments

3 comments

Sort by
  • priyasinha
    Hi,

    Thanks for your post.

    I am trying to reproduce your issue but no luck.

    I have tried the following scenarios:

    - I stopped the SQL Server service, to generate connection issue and then started the service again.
    - Tried wrong user name and password.
    - I also tried stopping and starting the actual Monitored Entity so that it looses connection and re-connects when the machine comes up.

    In all the above scenario, I never received any "Log backup overdue" alert.

    So could you help us in reproducing the issue please? Any help would be appreciated.

    Also, apologies but I dont think I understand what do you mean by reference-only DB. Could you point me to some documentation about reference-only DB?

    Thanks,
    Priya
    priyasinha
    0
  • PDinCA
    Reference-only = DB with "Numbers" table, "Dates" table", i.e., it has NO transactional implications whatsoever.

    I cannot reproduce the problem because Rackspace is responsible for the firewall configuration and I have no idea what they did. Whatever action they took, it caused SQL Monitor to stop monitoring and I had to re-apply the Windows credentials to the overall server row on the configuration page.

    The SQL Server service was untouched throughout the FIREWALL re-configuration. We were experiencing SSPI Handshake errors and Rackspace determined that it was "related to default lifetime of the pinholes used by our firewalls of the domain controller." I doubt you want to mess with your DNS firewalls; however, that's what caused SQL Monitor to quit and upon restart immediately emitted the Backup Alerts. I certainly won't be asking Rackspace to "put things back so we can provoke a SQL Monitor problem" :)

    Sorry I cannot reproduce for you. Are there any logs or the would the Monitor database itself be of any use?
    PDinCA
    0
  • priyasinha
    Hi,

    Thanks a lot for the information.

    I will try to investigate this further and would let you know if we can reproduce this issue.

    Regards,
    Priya
    priyasinha
    0

Add comment

Please sign in to leave a comment.