HealthCheck - Identify print jobs stuck in PRINTING state for suspicously long

Summary

In a rare situation the print job may get stuck in PRINTING state. This typically happens when:

• printer is supposed to confirm the job reception via LPR / IPP protocol related messages but it is not doing it

• printer buffer is full and there is a TCP Zero Window for considerably long time (e.g. in event of paper jam that is ignored by end user), this affects RAW protocol too

The user is usually (but not always, this depends on several factors such as the phase of job transfer when the printer stopped responding or the printer design) aware that there is some issue with his print job by looking into job list of the printer or by paying attention to error messages presented on the printer. Nevertheless users tend to ignore/skip/miss such warnings and information. In addition the job in PRINTING state blocks other job deliveries to the same printer. Any job queued up later on the same printer by any user ends up in a PENDING state, waiting for the original job to be delivered or for delivery to be terminated for example by MFD reboot.

In general it means that jobs stuck in PRINTING state for considerably long may have various negative impacts on a user experience:

• the job in PRINTING / PENDING state is not accessible/visible by end user on the terminal or in End User Interface

  • this is security measure to prevent user from releasing the job multiple times by mistake

User usually wonders what happened with the print job and then he goes back to a workstation, sends a job for print again and releases it on a different printer, but he is not happy about it.

Resolution

To minimize occurrence of jobs stuck in PENDING / PRINTING state kindly make sure you are using YSoft SafeQ 6 Build 67 or newer (defect fix SBT-3648 for IPP and IPPS). After the update there may still be a few occurrences because the root cause is outside of YSoft SafeQ, but the frequency should drastically drop since we improved the handling of limiting situations by implementing timeouts ippPrinterConnectionTimeoutSeconds and ippPrinterResponseTimeoutSeconds . Once timeouts expire the print jobs should switch automatically to Printer Error state and printer will not be blocked for everyone and the affected user can release affected jobs from the waiting list again.

If the issue still occurs, these manual steps will help to clear the stuck session:

Option 1

Restart the YSoft SafeQ server. This is quick solution but it has multiple downsides:

• Restart will terminate all active sessions, any user working on a terminal connected to the server or printing through that server will be affected and will have to try again once the restart is completed.

• Jobs whose delivery was terminated by restart may end up in CANCELED state ( CANCELLED state is not visible to end user on the terminal, administrator can re-queue it to secure queue to make it visible )

  • On very old releases around build 40 the jobs terminated by restart may remain in PENDING state until deleted by system parameter "maxSpoolerJobTime".

• If the server is member of a SPOC group a specific restart procedure needs to be followed:
  https://portal.ysoft.com/documentation-and-knowledge-base/documentation?page=How_to_Restart_a_YSoft_SafeQ_Environment.html

It might be actually better to leave the restart for the evening once there is less users using the system. The printer affected will be unavailable for printing for any user until then.

Option 2

1. Identify print jobs that are stuck in PRINTING state for more than 30 minutes

   1. First option - YSoft SafeQ Management Interface

      • set up a filter for print jobs that are in Printing... state

        

      • go through the list of jobs and review those where date is long in the past

      • you may review job details to see its size and think if it is so big for the delivery to take several minutes

   2. Second option - SQL query

      1. the query is to be launched on SQDB6 database (default name)

         MS SQL

         SQL

         select sj.cur_status_time last_status_change, sj.accept_time, sj.id job_id, sj.job_guid, sj.filesize /1024 job_size_KB, ('PRINTING') status, sj.cur_status_origin status_origin, u.login username, d.name MFDname, d.description MFDdescription FROM tenant_1.smartq_jobs sj
         left join tenant_1.users u on u.id = sj.user_id
         left join tenant_1.smartq_jobs_log sjl on sjl.job_id = sj.id and sjl.status = sj.cur_status
         left join tenant_1.devices d on d.id = sjl.device_id
         where sj.cur_status in (4)
         and sj.cur_status_time < (SELECT DATEADD(mi,-30,GETDATE()))
         order by sj.cur_status_time desc
         
         

         PostgreSQL

         SQL

         select sj.cur_status_time last_status_change, sj.accept_time, sj.id job_id, sj.job_guid, sj.filesize /1024 job_size_KB, ('PRINTING') status, sj.cur_status_origin status_origin, u.login username, d.name MFDname, d.description MFDdescription FROM tenant_1.smartq_jobs sj
         left join tenant_1.users u on u.id = sj.user_id
         left join tenant_1.smartq_jobs_log sjl on sjl.job_id = sj.id and sjl.status = sj.cur_status
         left join tenant_1.devices d on d.id = sjl.device_id
         where sj.cur_status in (4)
         and sj.cur_status_time < (select now() - interval '30' minutes)
         order by sj.cur_status_time desc
         

      2. the output example

         

2. Resolve the situation

   • Once you identified that job is really problematic, do this:

     • List all the jobs that are PENDING for problematic printer via YSoft SafeQ Management Interface > Reports > Job list

       • If there are such jobs, prevent MFD from accepting further jobs by one of the following:

         1. disable IPP/RAW/LPR protocols on the printer (via printer administrative website)

         2. block all the communication to printer via Firewall

         3. turn the printer off and keep it off until you finish all the steps

       • Note: This will prevent queued pending jobs from being released right away after clearing up the stuck job, this is meant as prevention for possible security incident where user no longer stands next to the printer.

       • Note: Action "Cancel selected jobs" on YSoft SafeQ Management Interface would have no effect, it is not meant to terminate PENDING jobs.

     • Terminate the delivery of PRINTING job

       • Download TCPView and run it on the PC from which the FlexiSpooler tries to deliver job to the MFD

       • Use View > Update speed > Pause (to prevent endless refreshing of screen)

       • Use View > Refresh now

       • Wait till all the connections are listed

       • Identify the TCP connection that exists between FlexiSpooler.exe and printer IP address > right-click it and choose Close connection

         • at this point job that was PRINTING is in PRINTER ERROR or CANCELED state (CANCELLED state is not visible to end user on the terminal, administrator can re-queue it to secure queue to make it visible)

       • Note: Using TCPView is preferred to rebooting of MFD, rebooting printer might not resolve the situation in combination of some vendor / delivery backend .

     • Double-check that connection is closed (e.g. refresh TCPView or use "netstat -abno <printerIP>")

     • Wait till all the jobs for the problematic printer get switched from PENDING to PRINTER ERROR or any other state

     • Reboot the printer, resolve any error message the printer is giving and enable IPP/RAW/LPR protocols or remove temporary Firewall rule

     • Verify that printing / accounting on the printer now work

3. Instruct users

   • to tend to an error on printer screen (if any) before leaving the printer

   • to initiate print/copy/scan only at the printer that does not report any error on its screen