Thanks, but that still doesn't get around the issue of checks that never change status, or how long they have been in their current status.
Mike >>> [EMAIL PROTECTED] 11/06/03 02:15PM >>> >From a post long ago, far away.. (see attachement) ________________________ Roland Gaspar Tel.: (514) 398-7880 xt 8603 ________________________ Roland Gaspar Tel.: (514) 398-7880 xt 8603 ________________________ Roland Gaspar Tel.: (514) 398-7880 xt 8603 >>> [EMAIL PROTECTED] 10/24/03 1:53:19 PM >>> Well, what if you saved and reloaded the same information you're writing to the table now? I'm not sure how you calculate that data internally. That would allow for the cumulative totals to keep increasing. But that would hide the fact that there was a period of time that the checks weren't done and is therefore unknown. If SA were down for a day or two, the %'s would be vastly affected. Now that we've been chatting about it, I'm not sure that changing the Statistics table is the right way to go, since what I need is pretty different than what is currently being processed. I think I actually want the new table shown below. Come to think of it, I don't really need the actual minutes spent in each state, just the xx.xxx% of total time for each datetime period. Can I request a new type of output table? Mike >>> [EMAIL PROTECTED] 10/24/03 12:30PM >>> Again that's not the question.... What exactly should be kept of the stats over a shutdown of SA.... dirk. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Michael Shook Sent: Fri Oct 24 4:39 PM To: [EMAIL PROTECTED] Subject: RE: [SA-list] Statistical database question.... Maybe a new database table with output set on it's own semi-daily schedule and on shutdown? Host | HostID | DateTimeRangeStart | DateTimeRangeEnd | Status 1 minutes | status 2 minutes | status 3 minutes | Status 4 minutes | status 5 minutes | status 6 minutes | status 7 minutes That's ultimately the table I'd need to create from existing data. Mike >>> [EMAIL PROTECTED] 10/24/03 09:48AM >>> The only time SA misses data is if the stats changes while SAlive is shut down. I do agree it is a problem if devices don't ever go down, they don't get an entry but you can then assume that it is 100% up. It would be nice to keep track of the number of hosts checked each cycle or to keep an average host entry count...I created a perl script that can run in ASP to parse the stat log and just have a stat at the bottom that anything not listed was at 100%. I have been playing around with the idea that at the end of each month, I automagically generate a copy of the current hosts file for Servers Alive, so up to a granularity of a month, I can list all the servers checked for that month and include it is my output. I wonder if dirk, would consider adding a feature that would append the list of hosts in the SAlive host file to the monthly stat log when he rotates the logs monthly? Dirk? > -----Original Message----- > From: Michael Shook [mailto:[EMAIL PROTECTED] > Sent: Friday, October 24, 2003 8:32 AM > To: [EMAIL PROTECTED] > Subject: RE: [SA-list] Statistical database question.... > > Well, the columns you currently write are fine, BUT.... > > I have been tasked to create monthly / annual UP/DOWN stats for the > major devices. What I do now is sum all of the minutes where the > previousstatus is X. The problem comes in that SA doesn't regularly > output state stats to this table. I only see entries if there has been > a change in status. So, unless my network is a complete piece of > trash, most of the devices never get any entries. Also, like I said, a > close/open of SA resets all of SA's internal time stamps so that if a > change does occur, the output only reflects from when SA was last > restarted. > > If SA output stats during a graceful shutdown, and on a regular time > schedule (say a few times a day (midnight, noon)) then I can use your > table as is. Otherwise, I need to perform the metrics calculations on > the interchange table, which will be a lengthy process if they ask me > for the last years stats, and I have been writing every 5 minutes. > > I can't see any issue with a previous and current status entry being > the same, but maybe a new column to explain more exactly why the > record was written would be helpful. > > Mike > > >>> [EMAIL PROTECTED] 10/24/03 06:04AM >>> > What exact info should be kept? > > > > dirk. > > > > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Michael Shook > Sent: Thu Oct 23 10:40 PM > To: [EMAIL PROTECTED] > Subject: [SA-list] Statistical database question.... > > > If I understand the process correctly, when the first cycle is run on > the start of SA you gather a status and datetime. When the status > changes, the duration is calculated, the datetime is reset, the new > status is recorded. Then an appropriate entry is written to the > database. > > If that's correct, then when SA is stopped/restarted, all of the > "running time" info is lost from the last program start or last status > change. > > Would it be possible to output all of the current "running time" info > for each check to the database before SA shuts down (assuming a > gracefull shutdown)? Maybe record the current status as the same for > both last and new. > > I'm trying to get the most out of this database, but I can't calculate > an accurate quality % if I don't have all of the up times. > > Thanks!!!! > > > > > > > > > > > > > Michael Shook > Business Continuity Analyst > Saddle Creek Corporation > 723 Joe Tamplin Industrial Blvd > Macon GA 31217 > 478 742 8740 ext. 105 (work) > 478 256 9318 (mobile) > 478 742 7917 (fax) > [EMAIL PROTECTED] > http://www.saddlecrk.com > > To unsubscribe from a list, send a mail message to > [EMAIL PROTECTED] > With the following in the body of the message: > unsubscribe SAlive > > > > To unsubscribe from a list, send a mail message to > [EMAIL PROTECTED] > With the following in the body of the message: > unsubscribe SAlive > > To unsubscribe from a list, send a mail message to > [EMAIL PROTECTED] c0 > Sent: Friday, Oct > With the following in the body of the message: > unsubscribe SAlive This email and its contents may be confidential. If it is and you are not the intended recipient, please do not disclose or use the information within this email or its attachments. If you have received this email in error, please delete it immediately. Thank you. To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message: unsubscribe SAlive To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message: unsubscribe SAlive To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message:tm when SA was last > restarted. unsubscribe SAlive To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message: unsubscribe SAlive xxxxxxxx Ce message a TtT analysT par Guinevere2 et McAfee This message was Scanned by Guinevere2 and McAfee http://nt.patlq.com/guinevere Questions? [EMAIL PROTECTED] xxxxxxxx ZIP attachment type(s) blocked *** This message contained attachments that have been blocked by Guinevere. Please see your system administrator for more details. Ce message contenait un (ou des) attachements qui ont TtT bloquT par Guinevere. Pour plus de dTtails, contactez votre administrateur de rTseaux. Questions? [EMAIL PROTECTED] *** To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message: unsubscribe SAlive To unsubscribe from a list, send a mail message to [EMAIL PROTECTED] With the following in the body of the message: unsubscribe SAlive
