What ticket status should I submit?

[Ron]

I'm just wondering:

What ticket status should I have submitted for a server having a load average of 20 and a one minute process taking more than 16 minutes to complete?

Level 1, 2 or 3?

What category should I put it in?
Server down?
Other?

What level should I select?
Low, Medium or High?

Oh and why doesn't any monitoring tool pick this up automatically? Why am I forced to be the monitoring script for this server?
sar output:

09:10:02 AM       all     40.13      0.98     12.19     32.69     14.00
09:20:01 AM       all     45.45      1.71     17.13     11.87     23.84
09:30:05 AM       all     58.05      1.14     22.98      6.88     10.94
09:40:02 AM       all     50.91      2.58     25.29      8.37     12.85

09:40:02 AM       CPU     %user     %nice   %system   %iowait     %idle
09:50:01 AM       all     48.45      1.45     21.51      7.25     21.34
10:00:01 AM       all     38.23      2.53     15.31     11.78     32.14
10:10:01 AM       all     37.22      0.41     18.02      9.38     34.97
10:20:01 AM       all     36.77      0.20     42.57      6.46     13.99
10:30:01 AM       all     35.45      0.91     43.61      8.46     11.58
10:40:02 AM       all     39.78      0.03     46.00     11.02      3.18
10:50:03 AM       all     44.83      0.01     46.29      4.16      4.71
11:00:02 AM       all     34.37      0.01     59.98      1.34      4.30
11:10:01 AM       all     38.21      0.02     59.88      0.58      1.30
11:20:04 AM       all     38.41      0.02     59.55      0.50      1.52
11:30:01 AM       all     37.62      0.04     61.92      0.14      0.29
11:40:01 AM       all     40.03      0.01     59.96      0.00      0.00
11:50:01 AM       all     41.55      0.02     58.43      0.00      0.00
12:00:02 PM       all     38.07      0.01     61.92      0.00      0.00

Ticket has been submitted (13469284)

[JPC-Zachary]

I'm just wondering:

What ticket status should I have submitted for a server having a load average of 20 and a one minute process taking more than 16 minutes to complete?

Level 1, 2 or 3?

What category should I put it in?
Server down?
Other?

What level should I select?
Low, Medium or High?

Oh and why doesn't any monitoring tool pick this up automatically? Why am I forced to be the monitoring script for this server?
sar output:

09:10:02 AM       all     40.13      0.98     12.19     32.69     14.00
09:20:01 AM       all     45.45      1.71     17.13     11.87     23.84
09:30:05 AM       all     58.05      1.14     22.98      6.88     10.94
09:40:02 AM       all     50.91      2.58     25.29      8.37     12.85

09:40:02 AM       CPU     %user     %nice   %system   %iowait     %idle
09:50:01 AM       all     48.45      1.45     21.51      7.25     21.34
10:00:01 AM       all     38.23      2.53     15.31     11.78     32.14
10:10:01 AM       all     37.22      0.41     18.02      9.38     34.97
10:20:01 AM       all     36.77      0.20     42.57      6.46     13.99
10:30:01 AM       all     35.45      0.91     43.61      8.46     11.58
10:40:02 AM       all     39.78      0.03     46.00     11.02      3.18
10:50:03 AM       all     44.83      0.01     46.29      4.16      4.71
11:00:02 AM       all     34.37      0.01     59.98      1.34      4.30
11:10:01 AM       all     38.21      0.02     59.88      0.58      1.30
11:20:04 AM       all     38.41      0.02     59.55      0.50      1.52
11:30:01 AM       all     37.62      0.04     61.92      0.14      0.29
11:40:01 AM       all     40.03      0.01     59.96      0.00      0.00
11:50:01 AM       all     41.55      0.02     58.43      0.00      0.00
12:00:02 PM       all     38.07      0.01     61.92      0.00      0.00

Ticket has been submitted (13469284)

Thank you Ron. Chris is looking at it. I'm thinking that is going to be a level 3 issue 100% of the time.

[Ron]

There doesn't seem to be a category for server performance or swamped server or whatever.

[JPC-Zachary]

There doesn't seem to be a category for server performance or swamped server or whatever.

The server is back to normal.

[Ron]

Yes, that was pretty quick. Thanks Chris (and Zach).

What category should that go into? Other? Server down?

I want to help properly categorize the problems you encounter so that the metrics can help you focus on where there may be issues.

There doesn't seem to be a category for server performance, so you guys must think that's never an issue?

[JPC-Zachary]

Yes, that was pretty quick. Thanks Chris (and Zach).

What category should that go into? Other? Server down?

I want to help properly categorize the problems you encounter so that the metrics can help you focus on where there may be issues.

There doesn't seem to be a category for server performance, so you guys must think that's never an issue?

Masood has added a "Server Performance" category for this types of issues. They normally would go in "Other"

[Ron]

Masood has added a "Server Performance" category for this types of issues. They normally would go in "Other"

I guess you meant "hasn't"

Well that makes sense now when he told me that the noise level is too high in the system to be able to get good metrics from the ticket system.

Greg, it wouldn't be for the benefit of the Rewards program, it would be for the benefit of creating the metrics so that you know what is causing me (and others, perhaps) to need to open the tickets. Perhaps it might bring the number down to 9,000 a month?

Is my experience that far removed from all other resellers? Or maybe the reality is you'd just be better off without demanding customers like me, or getting me onto my own server so that I can keep it running smoothly with nobody to ask for help from but myself.

as Ron points out, the number of tickets for actual JagPC side problems is something you should WANT to track anyway for much more important reasons. Do you really not make any distinction between playing helpdesk and getting actual requests? Then how do you measure your service quality internally?

Support tickets is not a metric for service quality. It can never be because its too variable. We have other internal mechanism to measure service quality. We have just started deploying further quality control mechanism and testing their results to control the very issues that Ron has reported in this thread.

WHATTTT?????

With 10K tickets per month? We MUST be talking about something completely different! That is a HUUUUGE sample.

You should be able, with good inputs on the tickets system on each ticket from the techs, what the problem was and what the root cause was and what the solution was. The easiest part to begin with is to identify if this was a customer-solvable issue (or a non-issue) or a "real" ticket identifying a real problem. From there you can break things down into services and then into problems with the services.

Then over a month or two you can SEE where the biggest problems still lie and develop automated solutions for them. If there are lots of uncategorized tickets you can look to see if they come from certain techs (retraining needed) or look to see if there are areas for which categories need to be created.

With proper ticket metrics you will see patterns emerging, you will see new problems emerging, you will see overloaded machines, you can detect hardware patterns, and you can check the workload and effectiveness of the techs and on and on.

C'mon, you're a big enough operation for stuff like this. No?

WHATTTT?????

With 10K tickets per month? We MUST be talking about something completely different! That is a HUUUUGE sample.

Exactly. The noise level is too high to measure service quality from it.

Omg...

By the way that conversation was just over three years ago. Have those quality control mechanisms been implemented? Are we catching servers with 0 idle time?

[Gwaihir]

They normally would go in "Other"

Interesting that you'd expect this. I put those in "server down", as that's the nett effect and these need the same allways-on-the-double attention. "Other" sounds way too leisurely to me for this sort of thing and to be honest: "server performance" does too.