Jump to content

DarrenWhite99

Administrator
  • Content Count

    1232
  • Joined

  • Last visited

  • Days Won

    166

DarrenWhite99 last won the day on April 1

DarrenWhite99 had the most liked content!

Community Reputation

410 Excellent

My Information

  • Location
    Redding, California, US
  • Agent Count
    2000 - 3000 Agents

Converted

  • OCCUPATION
    Senior Systems Engineer

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Fix heartbeat: https://gavsto.com/agent-response-slow-tired-of-waiting-to-interact-with-agents-offline-server-alerts-flaky-your-heartbeat-may-be-broken/ Enable Fasttalk in the script when it starts.
  2. Try \/[a-z]* ? But actually it’s best if you can post a sample of the output you are working with.
  3. Yes, it is in the database. I was commenting on the agent not sending that as a distinct value, but rather simply reporting who IS logged in and the server copies to the “lastuser” column. (There is a column for current logged in users, and one for the last logged in user. The agent updates the first column, the server updates the second)
  4. I don’t think the agent sends a “last user”. I believe it reports the logged in users, and each time it changes the database agent copies a name to that column. Logged In users are reported (at least) during each check- in, which is every 5 minutes normally. (Or every 30 seconds for probes or masters?)
  5. Umm.... Maybe it’s because I didn’t re-read the whole topic, but what problem are you asking about?
  6. That’s the problem. You could try replacing the TO_BAS64/FROM_BASE64 calls with BASE64_ENCODE/BASE64_DECODE. Those are stored procedures available in Automate, but they aren’t as fast as the native functions I’m using in 5.6+.
  7. It showed that there were no prior backups and nothing found to backup, so the SQL queries failed to return any results. Are you running MySQL 5.7? 5.6 (or was it 5.5?) doesn’t include the BASE64 functions needed so that could be the problem.
  8. I have an “onboarding accelerator” monitor. It triggers the onboarding script when it sees: A computer, online, in a location with onboarding enabled where onboarding has not completed and excluded from contract is not selected. I specifically use the v_extradata tables to test the EDF values because as soon as I can see the result there, I know that groups and such have rebuilt. (I tried triggering it quicker, but the standard onboarding script relies on groups/EDFs/etc. being up to date, so running it early means it misses things). That doesn’t speed the group builds, but it does start onboarding within about 20 minutes instead of only running once an hour. So, is the main issue that users are granted access to machines by specific groups, and the machines are not visible to them until the computer is in that group? A potential solution could be: Create a group that determines which agents they can see. Say you are using “Service Plans” as the group. You could have a search/group for “Agents Under Contract” or whatever, that just grabs all computers where the location has a server or workstation (as appropriate) service plan assigned with onboarding enabled and Exclude From Contract is not checked. Basically, it will be all computers that should be under contract. Ok, now, make a monitor that runs every 5 minutes to check if any not onboarded systems matched those conditions.. Trigger a script. (I think you already had all this design in mind) but instead of rebuilding EVERYTHING, just add a row to searches, subgroups and subgroupswchildren for the agent, adding the agent to the specific group(s) you need them in for technicians to access them. This would scale because you are only performing minor insert queries instead of asking for everything to be rebuilt for only one new agent. Not everything would be complete, but you could get them the access they need. (Oh yeah, also, some users permissions table would need to be touched..)
  9. It’s possible yes, but II don’t think it’s that great of an idea. ESPECIALLY if you are new to Automate, you wouldn’t know what problems can be caused or how to gracefully coordinate things. I’m thinking for example, what will your solution do when you are onboarding a client and 3-33 new machines get registered within a short window. Will your solution end up trying to rebuild groups and searches 30 times? Will that break all your internal monitors if they timeout while tables are locked? I’ll be honest and say that I don’t know all the trouble it would cause, but as someone who has used Automate for about 7 years and is familiar with many internal details at a level matched by few outside of the developers, my recommendation would be to reconsider your approach. Forcing these group and search refreshes are not things that you should need to routinely do.
  10. The script text for Execute Script must be ANSI. Is the script sending the email itself, or returning the email body and Automate is sending it?
×
×
  • Create New...