Passwords and Resets
Where eConnect holds a password, it is set on the user record and must satisfy the policy configured under Security Policy.
Resetting somebody's password
From their record, set Password and Confirm Password, and save. Give them the new password through a channel that is not the account you just reset, and have them change it at first sign-in.
The policy it must satisfy
Password rules are set system-wide and can be turned off entirely. When enabled they can require:
| Rule | Effect |
|---|---|
| Minimum Complexity | Which character classes must appear |
| Minimum Length | Shortest acceptable password |
| Lockout After | Failed attempts before the account locks |
| Lockout Duration | How long a lock lasts |
| Password Expires After | Days before a password must be changed; 0 never expires |
The full detail is in Security Policy. A password that does not satisfy the policy is rejected when you try to set it, so if a reset will not save, the policy is why.
Lockout
Repeated failed sign-ins lock an account for the configured duration. A lock clears itself when that time passes.
Lockouts are worth looking at rather than merely clearing. A user locked once has forgotten their password; an account locked repeatedly, especially outside working hours, is worth investigating before it is reset. Their session history shows the attempts.
Expiry
Where the password policy sets an expiry, a person whose password has aged past it is challenged at sign-in: they are asked to set a new one before the application opens, rather than being refused or let in on an expired credential.
That makes expiry something people deal with themselves, at the moment it matters, instead of a support call.
Where passwords expire, users are required to change them on the schedule set. Setting expiry to zero disables it.
Modern guidance favours long passwords that do not expire over short ones that do, since forced frequent rotation tends to produce weaker, more predictable passwords. Whichever your organisation requires, apply it deliberately rather than by default.
Directory-authenticated accounts
Where an account authenticates against your directory service, eConnect holds no password and none of the above applies — the directory's own policy governs, and resets happen there.
This is worth confirming before spending time on an eConnect reset that will have no effect. Such accounts are marked in the user list.
Users changing their own
People can change their own password from their profile, which is preferable — a password an administrator set is a password an administrator knows.
A second step at sign-in
From 11.0.10, people can add multi-factor sign-in to their account — a code from an authenticator app or by email, asked for after the password — and a user group can require it. It works for directory accounts as well as for the passwords eConnect holds. Three things follow for anybody handling passwords:
- A password reset does not remove it. Somebody with multi-factor sign-in still needs their code after you reset their password.
- It has its own pause. Ten wrong codes in a row stop code entry on that account for fifteen minutes. That is separate from the password lockout above, and it clears itself.
- Every password change made in eConnect forgets the person's remembered devices, so their next sign-in on each asks for a code. A change made in your directory does not.
What an administrator can see and reset, and what to do when somebody loses their phone, is in Multi-Factor Sign-In for Administrators.
Password reset by email
People can reset a forgotten password themselves with a link sent by email, once an administrator switches it on under Security Policy. It needs a working email server and the server's address in its configuration.
Forgot password? on the sign-in screen asks for a user name or an email address and emails a reset link to the address in the matching account's details. What the person sees is in Signing In.
It reveals nothing. Whatever is typed, the answer is the same — If an account matches, we have sent an email. The link expires in 20 minutes. — so the screen cannot be used to find out whether an account or an address exists. Asking changes nothing on any account: the current password keeps working, no session ends, nothing is locked.
Who gets an email:
| Account | Gets |
|---|---|
| An account eConnect holds the password for, super users included | A link, to the email address in its details |
| An account that signs in through your directory | A note that eConnect cannot reset it, and no link |
| Several accounts sharing one address | One email listing each account with its own link (up to 20) |
| Disabled, deleted and service accounts, and accounts without an email address | Nothing — and the same answer on screen |
Using the link. It works once, for 20 minutes, and allows five attempts. The new password must meet the password policy and differ from the current one. An account with the authenticator app must also enter a current code from it, or a recovery code — because whoever can change the address in a user's details could otherwise redirect the link. An account whose only method is email codes needs only the link, which already proves that mailbox. A passkey is never asked for or accepted there.
After the reset, the person is signed out everywhere, any password lockout is cleared, and their remembered devices, unfinished sign-ins and every other reset link are cleared. They are emailed Your eConnect password was changed. The reset does not sign them in, and their next sign-in asks for their second step as usual. Their sign-in methods are not touched.
Its limits
| Limit | Value |
|---|---|
| Per email address, and per account | One email per 5 minutes, 3 per hour, 5 per day |
| Per network address asking | 5 requests per 15 minutes |
| Whole system | Reset emails per hour (whole system), 100 by default |
| One link | 20 minutes, works once, 5 attempts |
Every limit is silent: a request it drops gets the same answer as one that was sent. Two consequences are worth knowing:
- People who reach eConnect through one network address share the per-address limit — behind a shared internet connection, a proxy, or the Service Bridge.
- Because a request needs no proof, somebody can use up one account's reset emails for the day, or the whole system's for the hour. The person is delayed, never locked out: their password keeps working, links already sent stay valid, and you can still set a new password for them.
What is recorded
A completed reset is written to the audit trail as
USER '{name}' PASSWORD RESET BY EMAIL LINK ({n} SESSION(S) ENDED). Nobody is signed in when a link
is used, so it is recorded under the user's own name, and it also appears in their
session history as an entry with no duration. Requests, and the emails sent or
dropped, are written to the server log only — never what was typed, an address or a link.
What is recorded
Password changes are audited under the user administration type. The password itself is never recorded — the audit trail shows that a credential changed, who changed it and when, which is what a review needs. See Reading the Audit Trail.
Failed sign-ins appear in session history, which is where a pattern of attempts becomes visible.