| Title: | DECmcc user notes file. Does not replace IPMT. |
| Notice: | Use IPMT for problems. Newsletter location in note 6187 |
| Moderator: | TAEC::BEROUD |
| Created: | Mon Aug 21 1989 |
| Last Modified: | Wed Jun 04 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 6497 |
| Total number of notes: | 27359 |
Hello, I have noticed a different inacuracy with the PA module which most likely needs attention for v1.2 if not already fixed. An example would be if you request past statistics for an interval; and the polling failed during the START time frame, PA can produce exceedingly high stats. What lead to this finding was that I asked for stats from 8am to 5pm as a single duration. show node4 xxxx circuit yyy all statis,for start 8:0 every 9:0 until 17:0 - dur 9:0,in domain zzz The stats showed utilization of 140%. I then requested counters for the same time frame and what was returned was a counter from the night before, then counters from about 10am up to 5pm. The counters were missing because the historian process had stopped. But I would think PA should better recognize the situation and respond with an appropriate error. best regards, brad...
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1590.1 | PA ok, user inacurate | JETSAM::WOODCOCK | Tue Oct 08 1991 10:36 | 14 | |
For anyone interested: I'm glad to say this turned out to be a user problem and not a PA issue. There were enough counters for PA to operate correctly for the scope of time. The missing counters were just before the SOT and it just proved to be a misleading clue. Anwar's fine math skills showed us to have ~89kbs on the circuit for the interval. Our records showed us to have a 56k circuit, line speed was set to 56k, and our internal vendor verified their records for providing us with 56k. WRONG. After a little investigating it was proven to be a 128k. When was the last time you got more bandwidth than you thought you had :-). Good stuff cheap. best regards, brad... | |||||