PagerDuty on UserVoice

Provide Amazon Web Services CloudWatch direct integration.

2012-05-16T11:22:33-07:00

Andy suggested:
For so many AWS users, the built in CloudWatch is the easiest way to monitor a bunch of server and service parameters. You may also be able to integrate directly with simple notification services (SNS) for an email-free solution. This can be handled through the generic email now, but it's not very smart.

Detect URLs in the event details JSON and make them clickable

2012-05-13T04:54:59-07:00

Steve Kryskalla suggested:
Small enhancement, but would save a step of copy & pasting the URL. Here's a screenshot: http://i.imgur.com/sqqdq.png

Allow a user to mark themselves as on vacation to immediately skip them in notifications [updated]

2012-05-10T15:11:25-07:00

I recently went on vacation and had to change several services/escalations to remove myself so things would immediately go to the next person. So it would be great to add a check-box to the user profile where someone can just mark themselves on vacation so they are skipped immediately. If it's in the user profile then we don't have to edit several things.

esaito said:

Same here... one of the engineers will be on vacation next week, and there's no easy way to remove him from the rotation.

I can use the Override function, but that won't allow just skip to the next on-call person.

Plus we have schedules for primary and secondary on-call roles, so I'd need to work something out manually for the 2 different schedules.

Please just implement a "I'm on vacation from X to y date" feature.

Allow a schedule-type that allows members to sign up to periods and swap with each other [updated]

2012-05-05T10:33:10-07:00

... instead of a round-robin automatic-assignment. We have a process where we expect people in the support-pool to pick which weeks (we work in week-chunks) they want to be on-call for, then sign up to those autonomously. We also allow them to swap out individual days with each other as needed. We pay extra for being on-call, so would want to track who's been on-call for how many days in period x (and, which days, since the rates are different for different days of the week). It'd be great if pagerduty supported this (does it already?).

waqaspathan said:

i like

support webhooks as a notification method [updated]

2012-05-02T13:41:35-07:00

Webhooks (http://webhooks.pbworks.com/) as a notification method would allow for almost endless customization on the end-user side. For example, we'd like alerts to be delivered to Campfire - if you supported webhooks, PagerDuty wouldn't have to write any code to support Campfire, we'd just setup a little webapp that did that for us, and point PagerDuty at it. GitHub's open source services repository is a great example of how webhooks can allow extensibility of an application by it's end users with little-to-no work by you! http://github.com/github/github-services/tree/master

Liz Carr said:

Webhooks notification is indeed being worked on! No ETA yet, but we'll keep you updated.

-the PagerDuty Team

support webhooks as a notification method [updated]

2012-05-02T12:58:45-07:00

Tim Sharpe said:

Can you at least tell us if it's being worked on or ignored?

Personal ical calendar so that my google calendar only shows my shifts. [is now completed]

2012-05-02T11:06:16-07:00

Provide a personal ical calendar with only my on-call schedule so that my google calendar isn't full of everyone else's shifts. Currently when I view my google calendar I can see who is working, but in general I only need to see if I'm on call at a given time or not. Having everyone's shifts clutters my calendar, especially on my phone where the screen is very small.

Personal ical calendar so that my google calendar only shows my shifts. [updated]

2012-05-02T11:06:16-07:00

Liz Carr (admin) responded:

Hi requesters! You can access personal calendar feeds by following the instructions here: http://support.pagerduty.com/entries/20631727-ical-integration

- The PagerDuty Team

Include "Acknowledged On" column in incident reports [updated]

2012-04-25T17:32:24-07:00

Our support team is judged both on time-to-acknowledgement and time-to-resolution. Currently there is no easy way to get the acknowledgement time, although pagerduty does have the data (see the incident detail page).

Richard Wun said:

This is a great idea!

Include "Acknowledged On" column in incident reports

2012-04-25T16:16:23-07:00

Kris Feldmann suggested:
Our support team is judged both on time-to-acknowledgement and time-to-resolution. Currently there is no easy way to get the acknowledgement time, although pagerduty does have the data (see the incident detail page).

Allow modification of escalation policy through REST API

2012-04-25T00:05:41-07:00

Earth Ugat suggested:
We have set escalation policies that notifies person1, then person2, etc. We integrate the incident-forwarding into our system through your REST API. We would also like to be able to modify the people in that escalation policy through the API, instead of having to go to the pagerduty website every time one of those people calls in sick, or something.

Allow more logic in Escalation Policies [updated]

2012-04-20T13:13:34-07:00

In particular, I want each lower level to continue receiving alerts (based on their profile settings) until somebody ACKs the alert. The way it runs now, as the alert escalates past a person that person is 'let off the hook' and will receive no further alerts even if the event goes unACKed.

Alexander Berezovsky said:

It might be helpful to have an option to inform team or even organisation with preset template customized with message passed with SMS reply like:

0:all:Working on this problem, ETA 15min
0:lvl3:Check your email for update

support 24-hour time formats

2012-04-20T12:07:52-07:00

Ryan Biesemeyer suggested:
Especially useful if the user has their timezone in PD as UTC (to better match UTC-based server logs without resorting to date maths), these AM/PM things are nonsense.

Reporting: Add Account Administration Capabilities

2012-04-18T09:18:09-07:00

Dan Nguyen suggested:
It would be nice to have the ability to perform a dump of all user accounts and their settings.

clean up the Layer UI

2012-04-17T15:42:42-07:00

Chris Doherty suggested:
1. The help video shows the ability to re-order layers, but I found no clickable area where that was possible. 2. Deleted layers get archived for historical purposes: why can't I reactivate them? Also, having "Layer 1" followed by "Layer 3" is jarring. 3. From the Edit screen, it pops up the "Leave This Page" dialog whether I've made any changes or not.

Incident Pick-Up Feature

2012-04-17T13:30:43-07:00

Alex Ho suggested:
To better facility group alerting, i think if we can include a "pick-up" button to assign the incident to the user who logged in first (as themselves) and perhaps go as far as having the option of sending a auto-reponse email back to the group indicating that the incident has been picked up by that team member. This makes people aware that someone's on it. Right now, it does not seem that the "acknowledge" or "resolve" assigns the incident to the actual user who logged in to push the button.

Create an iPhone Application with Push Notifications [updated]

2012-04-13T07:37:22-07:00

A good idea for an iPhone app :-) Push Notificaitons as an alternative to SMS. Be able to resolve, escalate, etc. from application. http://github.com/PRX/apn_on_rails

Gabor said:

Any update on this ?
It can be a huge backlog :)

Allow changing notification methods by time of day. [updated]

2012-04-12T12:36:46-07:00

When I'm awake, notification via SMS suffices. But when I'm asleep, SMS has little chance of waking me, while a phone call has a good chance. I currently have SMS as my first contact priority, and phone if I don't respond to that quickly. But I'd be likely to respond even faster if I could specify "only even try SMS within these hours".

Dana Kashubeck said:

I have the opposite situation, but being able to specify a time of day for the notifications is still exactly what I need. There is no reason for me to be texted for issues during the day when I'm on the computer working and am paying attention to email alerts.

However, I'd like to have text alerts at night when I'm asleep, as that will wake me up.

Create an iPhone Application with Push Notifications [updated]

2012-04-12T12:25:27-07:00

A good idea for an iPhone app :-) Push Notificaitons as an alternative to SMS. Be able to resolve, escalate, etc. from application. http://github.com/PRX/apn_on_rails

Dana Kashubeck said:

A mobile app with push notifications would be WONDERFUL! Would help keep me from going over my texting allotment.

Android App for PagerDuty [updated]

2012-04-11T11:01:27-07:00

Would be a great feature for those with Smart Phone

Aaron Cutchin said:

A full-featured smart-phone-based monitoring app should be more of an integrated client of the monitoring system rather than just a fancy SMS receiver. What if your phone fails to receive the SMS messages?

Such a client should be bi-directional. I.e., it should monitor the monitoring system, and alert the bearer if it cannot reach it over a specified retry limit or timeout.

This way, the on-call admin is alerted if they inadvertently wander into an area without cell phone or data coverage, or if the monitoring system goes dark for any reason.

It should also scream bloody murder if the phone battery reaches a critical level, and perhaps notify another oncall admin. What if I lose my phone when oncall, and don't realize it for several hours?

I've experienced many issues with relying on SMS for pager coverage: wandering out of range, SMS network failures (you've all received a SMS alert a day or so late, I'm sure), phone failure, etc. A bi-directional, integrated client would solve a lot of these!

Surface notes on the graph

2012-04-09T17:15:06-07:00

Wren Weburg suggested:
The new reports are great. Beyond the initial version, I'd especially like to find incident notes without having to drill down to individual incidents. For example, being able to mouse over icons on the graph where there are incident notes would be cool, so I could recall what a particular incident spike was and how it was fixed. Others have already mentioned ideas like being able to search notes and also to see aggregates of Incident subjects to find hot spots, which would be great too. In other words, this is a great first step to insight into the data, now let's take it further!

A default filter for all lists

2012-04-05T15:20:10-07:00

Shane Presley suggested:
It is incredibly tedious to type in a search every tab I go to, and every time it is the same thing, all of my schedules, policies and services begins with the same prefix. I would like some way of specifying a search criteria once and it persisting across tabs until I choose to remove it. Thanks!

Integrate with AlertSite for uptime monitoring

2012-04-04T16:57:48-07:00

thnntn suggested:
AlertSite (like Pingdom) is an uptime monitoring site. They do provide an HTTP POST notification, but don't allow you to configure the post data. I'm left with email alerts. Would like better integration

escalate to the next person in the oncall schedule

2012-04-04T10:24:47-07:00

Brian Spolarich suggested:
We have an weekly oncall rotation with three levels of escalation. Our policy is that the second and third level escalations are just the next two people in the schedule. Right now this requires creating multiple schedules which can be difficult to manage, particularly if we need to make a change for some reason. It would be great if we could just define the escalation policy to say "if the primary doesn't answer, try the next person oncall for this schedule, and then the next".

support webhooks as a notification method [updated]

2012-04-01T12:44:52-07:00

Greg Albrecht said:

THe last update from PagerDuty was many years ago, any update on this feature?

Export reports to tab/CSV

2012-03-26T15:49:03-07:00

Prakash Janakiraman suggested:
From the reports page, would be nice to be able to export a list of all issues in CSV or TSV format to be able run reports in Excel or other offline systems. This would be a lot easier than you guys having to build full-on analytics, and would just require making available for download the raw data so that we could analyze it ourselves. Examples: A list of incidents with date opened and date resolved. * incident id * service * time opened * time closed A list of alerts * incident id * alert time * alert type (sms, email, phone) * assignee * service name From those two feeds, we could do all kinds of analysis and graphing.

Incident groups [updated]

2012-03-26T13:28:45-07:00

Alerts frequently come in batches. The UI makes it really hard to manage multiple related incidents. It would be nice if we could select multiple incidents and merge them / group them under a single incident. The audit trail of each individual incident would all show up on the same page.

Ozzie Sabina said:

Related to this, we always add notes to incidents that we resolve manually. If we have like 20 incidents that are all part of the same master incident, it's hugely painful to put notes in each incident. Usually the best we can do is put a note in one and remark that the others are related. That's far from optimal.

Notify on problem resolution [updated]

2012-03-22T09:51:13-07:00

Currently there is no notification when problems resolve themselves (especially if they auto-resolve via returning to an okay status) This could lead to scrambling if you miss an alert which has already fixed itself.

Aaron Loughary said:

Any plans to add this? This would be a nice feature to add for auto-resolving tickets with the pingdom service. I get an alert when an incident is triggered but no alert when its resolved. With pingdom, it would alert when a site was down and when it came back up and it does this to PagerDuty to auto resolve the ticket, however, PagerDuty doesn't pass the resolved status along.

Heartbeat Alerts - Alert when job DOESN'T run by a certain time

2012-03-14T08:10:20-07:00

Joshua Silver suggested:
AlertGrid has a great feature that will trigger a an alert when a job doesn't run by a certain time. For example, it expects backups to be completed by 1am and send a signal. If it doesn't receive a signal by 1am, it alerts. This could indicate a job that failed to run or the server being down, etc.

Android App for PagerDuty [updated]

2012-03-12T06:27:36-07:00

Would be a great feature for those with Smart Phone

Vojta Rocek said:

Need!

Show public holidays (and other events) on the on-call schedules view

2012-03-12T04:04:58-07:00

Pete Mounce suggested:
So that I can see whether the schedule clashes with * public-holidays (so I can check whether I a) want to be on call then, and b) get paid the increased rate my company pays for overtime there) * company-events (we have those) * private-calendar events and plan my on-call around those.

Integrate into Jaconda (or other external services)

2012-03-09T10:08:06-08:00

Peter Baker suggested:
Would be great if changes to calendar, alerts, etc could be integrated as a notification type in Pagerduty. https://github.com/jaconda/jaconda-api

Create a "Quick Override" button on the dashboard

2012-03-08T22:41:16-08:00

Geoff Galitz suggested:
This button would automatically create an overrride which would put the user on-call as an override for the next 8 hours. This would enable team members to cover for each other during any kind of emergency without having to dive through the different menus and figure out timezones during a urgent situation. Also, the button should change to "Cancel Quick Override" after it has activated for when the emergency is over.

Include a link to acknowledge an incident in alert emails

2012-03-08T03:01:53-08:00

Arto Bendiken suggested:
It is annoying that alert emails don't include a link to acknowledge the incident in question, in my case frequently requiring me to scramble to login to PagerDuty and load a relatively-heavy web page (very slow if I happen to be connected over a GPRS-rate mobile connection) just to prevent it from unnecessarily escalating to an SMS/phone alert.

Add more per-incident reporting [updated]

2012-03-06T23:08:35-08:00

Would be good to have more reporting showing the number of incidents over some time period to track the amount of alert activity. Ideally would be good to track that activity broken down by service, and even the message subject to see "hot spots" in the infrastructure.

Ozzie Sabina said:

Ditto. We'd actually like to be able to have a post-incident (or maybe even during incident) reporting mechanism. Tying this into the ability to add notes to incidents makes this a great all around central spot to keep people informed as to what's going on.

Create a sortable Oncall sidebar

2012-03-06T05:41:00-08:00

Thomas Vachon suggested:
We have multiple oncall groups, it would be best if you could set a sort order in which they show up. Also the scroll box inside the lists is very wierd and should just be gotten rid of.

Allow people to me carbon-copied on incidents without assigning the incident to them [updated]

2012-03-05T15:14:14-08:00

Our management would like to be paged on all incidents but without actually assigning the incident to them. Right now we simulate that by having a fake sure that just SMS all managers and assign the incident to that user for 1 minute before escalation. But that method slows down notification & is hard to maintain.

Joshua Miller said:

Sergey that does not work if you use the api, right now we have a dirty hack that also sends alerts to watcher group which is a pain make sure its in sync.

connectwise/autotask integration [updated]

2012-03-05T13:02:58-08:00

we use connectwise for managing our support tickets (packettrap for system monitoring). both have open api. it would be great to have integration

PagerDuty Support Team said:

Right now we have several customers using the default email integration with ConnectWise successfully.

In terms of a tighter integration though, I'll see if I can add ConnectWise to the list of code samples.

Android App for PagerDuty [updated]

2012-03-03T22:36:59-08:00

Would be a great feature for those with Smart Phone

Adam Strohl said:

Super needed as there are no good android SMS alert apps that can be tailored for specific patters (PagerDuty sends alerts from different numbers so you can't do per-contact). PLEASE DO THIS SOON.

Allow Zendesk Integration [updated]

2012-03-03T14:35:10-08:00

Right now we send an email from Zendesk to PagerDuty to open a ticket. But if we were to respond to the ticket, it would be nice for PagerDuty to resolve the issue on it's side.

Sarkis Varozian said:

We also need this, any efforts being made towards this?

enable MSN / SKYPE / ... instant messaging alerts [updated]

2012-03-02T12:38:15-08:00

We would like to be notified via MSN or SKYPE chat message when alarm occurs. The response chat message from us could then for instance signal that the alarm has been handled. --------------------------------------- Example: PagerDutyAlarmFriend: Alarm XZY was triggered. Please enter ok, to register your acknowledgement. User: ok -------------------------------------------- Thank you.

Ben Roux said:

Exact same sentiments. I feel that I am reachable via SMS or IM almost 100% of the time. Would be an excellent feature to add.

Warm Handoffs

2012-03-02T11:19:56-08:00

Thomas Vachon suggested:
For instance, user1 is oncall until noon. User2 must anwser a handoff confirmation at handoff minus 15 minutes before the handoff will occur. PagerDuty ensures that the user2 knows they are oncall

Folders or directories in PD? John Labbe Feb 29 20:36 As the amount of information in our pagerduty instance grows it becomes less managea

2012-03-01T21:03:00-08:00

John Labbe suggested:
Folders or directories in PD? John Labbe Feb 29 20:36 As the amount of information in our pagerduty instance grows it becomes less manageable. I often manage things now by using a prefix in the naming convention so they are at least grouped together when listed. Any ideas along the lines of directories or folders? Comments User photo Liz Carr PagerDuty Hi John, Thanks for the suggestion! This is not something that has been previously requested, as far as I'm aware. I would suggest adding it as a feature request on http://feedback.pagerduty.com/ and upvoting it. Thanks, Liz Liz Carr PagerDuty 650.989.2965 x 135 liz@pagerduty.com

Let me drag to nudge schedule overrides left and right rather than delete and re-add if I miss the slot precisely

2012-02-29T01:29:28-08:00

Pete Mounce suggested:
It'd be a nicer UX for overriding a schedule.

Allow a schedule-type that allows members to sign up to periods and swap with each other [updated]

2012-02-28T12:55:13-08:00

Liz Carr said:

Hi Pete. For the situation you describe, we'd suggest creating a basic schedule with one person on-call (perhaps a mailing list?). You'd then use overrides to allow anyone on the account to schedule themselves as needed. You can find out more about overrides here: http://support.pagerduty.com/entries/20565763-overrides If you have further questions, please send us an email at support@pagerduty.com

Allow a schedule-type that allows members to sign up to periods and swap with each other

2012-02-28T08:55:13-08:00

Pete Mounce suggested:
... instead of a round-robin automatic-assignment. We have a process where we expect people in the support-pool to pick which weeks (we work in week-chunks) they want to be on-call for, then sign up to those autonomously. We also allow them to swap out individual days with each other as needed. We pay extra for being on-call, so would want to track who's been on-call for how many days in period x (and, which days, since the rates are different for different days of the week). It'd be great if pagerduty supported this (does it already?).

allow incidents from a phone call (1.800.xxxx #123) [updated]

2012-02-27T14:58:16-08:00

Allow users / clients etc to call in to a phone number to create an alert. Ideally the recorded message would be stored as an attachment to the alert.

PagerDuty Support Team said:

The system doesn't support this directly, but you can script your own system to handle it through the PagerDuty API pretty easily. One example is here: http://blog.pagerduty.com/2012/02/triggering-an-alert-from-a-phone-call-code-sample/ that will take a phone call, record it and escalate it as normally with a PagerDuty alert. Smart phones can listen to it or call back from the SMS, but it won't play the MP3 over the phone.

Add comments to incidents [is now completed]

2012-02-24T22:05:51-08:00

It would be great if I could leave comments on incidents to remind myself and others what is going on. This is especially nice for acknowledged incidents that can't be resolved right away.

Build analytics (charts, graphs, reports) on events [is now under review]

2012-02-24T16:23:23-08:00

Analytics: metrics, stats, charts, reports on usage: - average response time: by person on call, by person not on call - escalation path (by day, week, month) e.g. issues for a specific service went through 3 people/hops until they were resolved - currently I don't see any counter per issue (e.g. 3 emails with exact content arrived), but there was no counter to tell you an issue had 1 trigger instance or 3 - #issues by hour - stacked graph of # issues, # acks, # resolutions per hour and per different services - badges for best ratio of issues to response for on call people (during X's on call time, the response time was 15min, during Y's time it was 24min) - chart with best on-call responses - chart with targets: I want the service to go from 50 triggers/day down to 2 triggers/day - i.e. we're getting swamped here - suggestions to admin: person Z is not very active/responsive - work on them and many more analytics & reports on this. Thanks for listening!

Create a daily report/Search for incident [is now planned]

2012-02-24T16:22:20-08:00

At a user-configurable time, send an email to the oncall (or, again, user-configurable target) with the previous day's incidents, including incident description, id, start time, resolve time and duration.