https://linen.dev logo
Gmail Polling Trigger no longer sending events
# 🖥│support-forum
c
This trigger is no longer posting events ti_WcgnDFXkkhf4 it is an active trigger. other triggers like it don't appear to be having problems
u
Hey @Chad Holmes, I can see a few payloads around 45 mins post your issue submission. Could you check and let me know?
c
Yeah some came through later but many were not temporarily and there was no reported issue
a
We are experiencing very similar issues with gmail triggers. I have created 4 support tickets. a. Triggers suddenly stop working for hours and will eventually fire all events at once. b. The new_gmail_message trigger not sending events, for outbound emails (config is correct). This is random for some triggers c. Triggers oAuth expiring arbitrarly and our users need to reconnect their Google Accounts d. Errors while switching/toggling the state of a trigger These errors are random and very hard to catch across our user base, but lately this has been crippling. Hope your team fix this issues soon @Uday
c
Yeah definitely problematic @Uday We have a connected account with Gmail not showing any received email triggers ca: ca_3GJSdUhC84be trigger: ti_kbVdM2OBqKKs Sent trigger appears to be working fine. Both show as active though so not sure why the other isn't working Here is the sent trigger ti_QOoYTDfGg2DQ
I have a specific email that we sent to the user during the time received emails are not creating events that we can verify against
u
Hey @Chad Holmes, we are investigating your trigger issues.
c
Hi there, any findings? We need an explanation for our user(s) @Uday
@Uday this is critical for us and is happening repeatedly. Each time we report this we are not getting answers. We will be moving away from Composio if we can't get insight into what is causing this issue and if it is being fixed
u
Hey @Chad Holmes, really sorry for the inconvenience. Let me check with the team and revert back as soon as possible.
c
Appreciate you responding @Uday Just so the team knows, we are scaling our user base and will be deciding if Composio can grow with us or not. This issue is a black box right now and I have no next steps for resolution for our user other than us solving the issue without using Composio. The tool is cool and worth the money if it works reliably and we can get support.
If this is a tier thing and we need to pay for priority support we will @Uday
For troubleshooting, it looks like a reconnection for the user did work but the main issue is how do we know when to prompt users to do this?
u
Hey @Chad Holmes,
ti_kbVdM2OBqKKs
seems to be working fine. — I can see payloads on this trigger. Could you please clarify whether the device was functioning normally before it suddenly stopped or if it never worked at all?
c
That trigger was removed. It stopped sending events before the outage began. The user reconnected their account which created new triggers and it began working again right before the outage then killed it again.
@Uday
u
@Chad Holmes, could you specify what you meant by "killed it again"?
c
the trigger was active. But events stopped
@Uday
u
Got it, we are checking it.
r
When I create a trigger and set the interval to 1 minute, it still gets changed to 15 minutes. After saving, when I reopen it, the interval shows 15 minutes instead of 1.
u
Hey @Raphael Kalandadze, I've escalated the issue with the team and they'll be investigating the issue.
@Raphael Kalandadze, we've changed the trigger interval behaviour and it is intended. The Dashboard should be updated soon as well.
r
So this means I can’t quickly trigger Gmail via a webhook if I have to wait 15 minutes for the webhook to arrive. How is that usable in production?
@Uday
c
@Uday if this is the case composio is no longer usable for us in production either.
@Uday Also this wasn't announced at all and now we have no time to prepare for this transition. Just silently changing the minimum behind the scenes is the worst possible rollout strategy I could think of
@Uday Can this be escalated? I would like to reach someone in leadership about this.
b
Hi, thanks for reporting and apologies on the outages and also on the sudden change in the interval field. We have been receiving a lot of traffic, requiring us to make some changes in the configurations to bring back up our triggers infrastructure. I do agree that suddenly seeing the interval at 15 minutes for new triggers is a bad experience and could have been pre-communicated. We are dedicating massive amount of resources onto this problem to not just scale and stabilize our infrastructure but also add a lot more new kinds of trigger slugs, as we are seeing a lot of demand. We will keep sharing more proactive updates on our changelog further on triggers. Please feel free to pick up some time from my [calendar](https://calendly.com/rahul-lingala-composio/new-meeting) to get more details and provide more feedback so that we solve this problem better for you.
c
I understand this is a scale problem. But why not 2 minutes or 3 minutes? 15 is a crazy adjustment. The trigger is useless to us at that point it is not a trigger. Now I am forced out of your platform with no opportunity to transition. We will be building the polling mechanism ourself. If this was your intention then that worked out great but I would've paid menaingful money for a reliable email trigger mechanism
b
Hi Chad, thanks for your reply. I have enabled 1 min polling triggers for your project (nanoId: Us8......)**.** Please pick a time on my [calendar](https://calendly.com/rahul-lingala-composio/new-meeting) so that I communicate our long-term plan to you in more detail and receive feedback at the same time.
c
@Lingala I am in the waiting room if you are still available
l
Hi I'm also in the room
Let me check my link
Hi @Chad Holmes , I responded about to your query on our Serious Business plan on email