Hi, On Fri, 12 Apr 2002 16:46:13 +0200, Alfredo Sola <alfredo@intelideas.com> wrote:
I fail to see all the factors contributing to the long queue. It's
Well, let me guess. I looked at http://www.ripe.net/ripencc/about/staff/hm-staff.html and saw 18 hostmasters there. Then I looked at http://www.ripe.net/ripencc/mem-services/registration/rttqueue/rttwaitqueue.... and saw that last 12 months there was average 189 tickets in the hostmaster queue. So, 189/18 = 10.5. That is, every hostmaster always works with 10.5 requests. Next, on the second page we see that during last 12 months the average request processing time is 9.3 days. Assume that every hostmaster processes his/her requests somehow parallelly or sequentially, and this takes 9.3 days. For one request this is 9.3/10.5 = 0.89 days. I suppose this value is the real average _total_ amount of time that every request requires to be completed. So, if we want our requests to take 3 days, a hostmaster sould have 3/0.89 = 3.37 requests. Under last year's average NCC work load, there should be 189/3.37 = 56 hostmasters. Significant? No more significant than 3 days against 10. Another digits. I've just opened Slides Booklet taken from last LIR training courses and saw that currently NCC has 3100+ LIRs and 99 staff members (slide 10, "Vital Statistics"). This is 31+ LIRs (read: address spaces, ASNs, DB objects, problems, questions etc) to one staff member. How do you find it? Stephen wrote that he doesn't see the lack of staffing but I do. If someone's resources are not enough for his load, he should increase the resources. But looking at that 12-months statistics we see that the load of hostmaster staff varies from season to season. Obviously we should understand that the request processing time is below average in autumn and winter and is above average in spring and summer. Colleagues, let's be patient, good things don't come fast. -- Aleksey A. Perov Postmaster ALP215-RIPE JSC Svyazinform, Penza, Russia e-mail: algardo@sura.ru phone: +7 8412 520215