Client behavior when server is down (0.6.1 and 0.7.0)

[airandfingers]

We're currently using ldapjs for our LDAP client, in secure mode:

ldap.createClient({
    url : 'ldaps://192.1.1.1'
    /* tls options if using ldaps */
    tlsOptions : { ... }, //key, cert, passphrase, rejectUnauthorized: false, NPNProtocols
    /* turn on connection pooling */
    maxConnections: 10,
    /* 2 minutes */
    maxIdleTime: 120000,
    /* 30 second health checks */
    checkInterval: 30000
});

When our client starts up, the first call it makes is client.search, passing a callback that takes err and res. If the server is down, this callback is being called twice, once with a "connect EHOSTUNREACH" err, and again with a "socket hang up" err.

I tried upgrading our ldapjs version to 0.7.0, but for that version we're seeing an even worse issue - the "connect EHOSTUNREACH" error is unhandled, so our entire app crashes.

events.js:72
        throw er; // Unhandled 'error' event
              ^
Error: connect EHOSTUNREACH
    at errnoException (net.js:901:11)
    at Object.afterConnect [as oncomplete] (net.js:892:19)

Is this normal ldapjs behavior, or are we using ldapjs in some non-standard way? All we want is to be able to handle errors like these, rather than working around double-callbacks or adding try-catch blocks around every client.search call.

[pfmooney]

See if pfmooney/node-ldapjs@6b7906c helps at all.

(edit: amended commit)

[airandfingers]

Yes, that worked. Thanks! Will this be in NPM soon, or should I use a cloned copy for now?

[pfmooney]

I would use a cloned copy for now.