While under no circumstances would we find it prudent to issue an update to our software the weekend before a week of celebration and relaxation for most of the world, we must announce the release of Matador Jobs 3.5.5. We strongly encourage you all to please update your sites immediately!
An Emergent Issue Discovered
Beginning yesterday, we saw a number of our sites begin to encounter failures when submitting applications to Bullhorn for candidates who already existed in their Bullhorn account. This was causing our routine to encounter a breaking error, resulting in all applications affected becoming “stuck” and, for sites with the “sync applications” option set to “submit applications immediately”, was causing a site-breaking error for applicants.
Beginning Thursday, December 18th, Bullhorn began rolling out the 2019-December updates to the ATS and a majority of the clusters got the update on Friday, December 19th, which was when the issue was first reported to us. We feel this is no coincidence, and may imply one of the following:
- We were not notified of a change to the API and we could not prepare our users for the update.
- The update caused an error on Bullhorn’s platform, and we must wait for them to fix it.
After hours of attempts, we were unable to determine a fix for the issue. We submitted various notice to Bullhorn techs requesting assistance. Because we suspect the holiday may cause this issue to persist for a while, we turned our attention to finding a work-around to restore the functionality of Matador in the interim.
Note: You Lost No Data
Prior to the first release of Matador, as told in our story, Paul and Jeremy ran precursor solutions that would lose applicant data when communication with Bullhorn failed. One of Matador’s biggest launch features was to store application data locally until a successful communication with Bullhorn completed. This ensured that no data would be lost.
This is true for this issue too! You lost no data!
With this release, we improved error handling for errors like these and notably gave Matador the ability to continue when a failure of this kind occurs. What does this mean?
- If Matador cannot update the existing candidate (updates usually are of form-submitted information like new phone numbers or emails), Matador now can skip the update altogether. This means the new phone or new email will not be changed, but can still be accessed in the resume and would be included in the recruiter email notification (if you have those turned on).
- Now that the process is able to continue, new files (resume, etc) will be uploaded to the candidate, and the candidate will be submitted to a job via a submission or web response depending on settings and if the application was connected to a job.
- Your next automatic sync with Bullhorn will scoop up and sync all the previously “stuck” applications, and if your settings remove completed applications, they will be removed as should be.
Next Steps Regarding Restoring Functionality
We have support tickets submitted to Bullhorn and expect a response at the start of the week. We will continue to monitor this situation and restore the intended behavior you’ve come to expect as soon as we are able to. Thank you for your understanding and your support.
Typically, we collect a small number of bugfixes and hold them for release when prudent. Included in this update were the following bugfixes already certified for release but that we intended to save for a post New Year’s update.
- Fixed an issue where attributes that inform the client-side form validator script were not included for form fields of the “select multiple” type. If your application contains form fields of that type and it has attributes defining it as required, this will now produce proper error messages for users trying to submit invalid forms.
- Fixed another issue around the handling of candidate skills as returned from the resume parser. When the failure was encountered, Matador continued, so no applications were held back as a result of this issue.