SamS

Hi @Rob Gething, The above code seems to be assigning a call. Are you sure that the analyst you are assigning the call to is allowed to have the call in their queue? IF the assignment fails, then you should get an error message back that sheds some light. Is your script showing the error message? If so, how about outputting assignError instead of the hard-coded text? If a call is correctly assigned to a team, but not to the correct analyst, then the Analyst ID must not be correct. The assignment, as a minimum, needs a correct call reference and team id - the analyst id is optional, so if you are not failing, then the analyst id is being overlooked for some reason.

Hi @Rob Gething, You might be more successful on a powershell script forum for this one. A Google Search on "powershell pass variable to another script" got me this promising solution (amongst others): https://stackoverflow.com/questions/26731933/how-do-i-pass-arguments-or-variables-from-one-powershell-script-to-another

Hi @Rob Gething, Webhook would be delivered on whatever you set the endpoint as. So, if the endpoint is https, then you should be fine. As for the error at hand: it looks as if iBridge is compaining that the account set up in KeySafe that you are using for this doesn't have the specific permissions necessary. Please confirm that the account exists and is not disabled in some way.

Hi @Martyn Houghton, The SLA Rules are worked through before the call is logged - effectively in the Service Manager interface - not after a call is logged (as part of the BP). The Request Importer script(s) do/es not use the interface, logging calls directly in Hornbill. It consequently does NOT know anything about any SLA rules. If you want to implement the SLA logic, then you will need to apply that logic in the request data source. Please note that the request importer scripts are made as one-off imports to get data into Hornbill - with a one-off effort at script configuration time. The script was not expected to be oft-used or replicate any interface logic (i.e. like Progressive Capture; if you want to map data, you need to manually map it).

Hi @AndyHill, You might find the latest release better - https://github.com/hornbill/goAzure2HUserImport/releases/latest Thanks for reporting the issue.

Hi @AndyHill, Are you trying to have two functionalities? i.e. removal of [] on WorkPhone and NO removal for Phone? If so, then the best thing would be to run the import twice - one with WorkPhone configured WITH the tweak set (NO mapping on Phone), and the other (without the tweak) mapping to Phone (but NOT on WorkPhone)

Hi @Martyn Houghton, It looks like you are configuring on the back of the configuration file alone. There are clarifications (in this particular instance the mention of "Codes") in the documentation: https://wiki.hornbill.com/index.php?title=Service_Manager_Request_Import_Utility#CategoryMapping

Hi @Paul Welby, Does anything stand out for you in the log file? Is something happening in the log file which you wouldn't expect to happen? Is something happening too often for no good reason (eg search for customer "")?

Hi @Paul Welby, I will actually be working on adding some features to the Asset Import script next week, so if your debugging highlights something, I'd be obliged. I'll have a look to check whether the logging near creation happens to mention updates to often.

Hi @Paul Welby, From which version did you upgrade? Did you re-use your existing configuration files? or did you merge in the new options that might have since been created? If an asset is already in the system, it should update the details if the OperationType is empty (default, in case you didn't set it specifically), set to "Both" or "Update". Otherwise it will create the asset - provided OperationType matches "Both" or "Create" or is empty (again the default if not set). A recent addition is to match users to their assets, this searches for each of the behind-the-scenes ID of each customer - although it does keep track of previously found users (so it should not make a new network connection for the second jbloggs it comes across. Adding the option "-debug" to the command line should give extra data in the log file to find out what is happening.

@AndyHill, The '''StringCollectionTweak''' in 2.7.0 should have automatically selected the first entry in the array - I will look into it - and fix it in v2.7.1 if I can replicate the issue.

Hi @AndyHill, Are you saying that telephoneNumber doesn't work for you anymore? I'm not entirely sure how Azure works behind the scenes (I think it has switched about a bit - telephoneNumber should return the primary number indicated in businessPhones and mobilePhone, the mobile number indicated in there, but I know telephoneNumber doesn't return anything if it is not expressly set), but I have noticed that telephoneNumber doesn't work for everyone - and for those where it doesn't work, the businessPhones route is now given as an alternative. You might also want to add businessPhones to the list of UserProperties.

@vhayer, I have modified your link above to highlight the part of the URL which happens to be the [[Instance ID]]

Hi @m.vandun, The scripts shouldn't be triggering it multiple times, my concern is that perhaps there is a proxy/firewall layer in between which (for whatever unaccounted for reason) might be throwing a spanner in the works. Are you running with "deleteReportInstance" set to true? if so, please set to to false, so we can confirm whether a report was triggered by the PowerBI. Another option would be to increase/set "suspendSeconds". Also, what are the results if you run a(n existing) "HistoricReport"?

Hi @m.vandun, Are you the only one triggering those reports? or are more (PowerBI) users/processes (eg a dashboard or so) potentially querying that same report at (nearly) the same time? Is there a load-balancer/proxy or something going out to Hornbill which might trigger the running of the report before your connection is made? I remember once dealing with IE which hit a WebServer twice on a POST: first time with an empty payload (to trigger an expected authentication) and the second time with the payload. If, in this circumstance, a middleman sends two full payloads (i.e. triggering the report to run on the first and the "report already running" on the second) , and only passes the second response to the client (R-script), then that might account for the above behaviour.

Hi @Frank Reay, Just to be on the safe side: did it work for BOTH (i.e. the cleaner AND the CSV Asset Import)?

Hi @Alberto M, The API Key should have the exact same permissions as the user for which the API key was generated. IF you log in as that user, can you make the data change you want to do via the interface? IF not, then provide the user with the necessary access permission (Role/Team membership etc) to make that access happen. i.e. if a regular user is not allowed to modify another user's profile, then your API key will not be able to update that either.

Hi @Frank Reay, No - no update. As you reported that the cleaner program now did not work, I assumed you realised that issue is environmental/network. On further inspection: it appears that you are using "Landmark" as your Instance ID in the command line, it will need to be "landmark" (all lowercase). If the cleaner program works, then we can rule out network issue, but if you also have "Landmark" as your instanceId in your CSV Asset import, then that will need lowercasing as well.

Hi @Frank Reay, I have just double-checked the release I uploaded and that isn't failing on my side. I would venture that the issue is thus environmental. If we have successfully excluded the network, then we are looking at either an issue with the configuration, the .csv file OR both. IF the configuration file is as you gave it above, then the obvious field missing is the h_name - to match the updates to the assets you are looking to update. Also, assume the headers/fieldnames of the .csv are case-sensitive (just in case it is "supplierid" in the source).

Hi @Frank Reay, I'll look into the particular issue you report, but the latest release is not relevant to your particular issue (hence no note of that in my previous response here).

Hi @Frank Reay, Can we confirm there are no other (network; eg firewall(*)/proxy server(**)) related configuration which might be interfering with the network traffic? Trying to obtain the asset class is the first thing the script is trying to do and it is communicating with Hornbill to do so. (*) the list of servers to whitelist is the same as for our other scripts (**) If you are using a proxy server, same as for other scripts, have you set HTTP_PROXY?

Hi @Brendan, Your query is incorrect, you are assigning "value" to disp. As the field names is already "disp", you can just suffice with : SELECT value as val, disp FROM SomeList.csv You can also simply "select value, disp from SomeList.csv" and code in "value" instead of "val". You might also want to change the {.disp} to [disp] as well (in the Translations section). I spot that what you have is part of the sample config we deliver, I'll have a look at tweaking those.

Hi @samwoo, Would an "Action" selection of "Create", "Update", "Both" be of use to you? i.e. you could then have one configuration file to use for creating assets (i.e. when the script doesn't find an asset to update and hence creates the asset), and another file to be used for updating existing assets (and not bother importing new items). "Both" would be the default/current behaviour of creating when not found, update otherwise. If so, I can likely add that functionality by the end of this week.

Hi @lee mcdermott At the moment we don't have any particular CTI integration. Your CTI solution MAY have capabilities to trigger the running of scripts based on certain actions (i.e. picking up the call, ending the call). Through that it should be possible to communicate to Hornbill. WHAT can be communicated will ultimately decide what can be done to marry Hornbill with your particular CTI platform. We appreciate CTI is a standard, and on the basis of that, we will likely be looking into a more robust/integrated solution at some stage in the future.

Hi @RobW and @Adam Higgins, Our developers have been made aware of this and are on the case.