function readOnly(count){ }
Sign Up ›
Starting November 20, the site will be set to read-only. On December 4, 2023,
forum discussions will move to the Trailblazer Community.
Learn More

Browse by Topic
ShowAll Questionssorted byDate Posted
+ Start a Discussion
sdetweilsdetweil 

inconsistant execution with Workbench

using the workbench I see a connection failure after each or every other account update or create, it does not recover.

 

Connection Broken
successful=false
failure=true
channel=/meta/connect
request.channel=/meta/connect
request.connectionType=long-polling
request.id=43
request.clientId=f1kszovhvnv9ufibfe7qn0ez1
advice.reconnect=retry
advice.interval=0
action=connect

altaiojokaltaiojok

I had one other person report similar behavior, and it seemed to happening for some orgs and not others, and possibily instance-specific. Can you let me know which instance your org is on (e.g. na1.salesforce.com, cs3.salesforce.com, etc) and I can try to look into it deeper.

 

Thanks,

Ryan
altaiojokaltaiojok

A couple more questions:
1. Does this happen only when a push topic event (i.e. create/update of account) happens, or do you also see problems with the initial handshake and connection establishment? In the other example I saw, the user was having problems getting the handshake and connection to be established successfully and the subscription weren't registering correctly because the connection kept dropping.
2. Rule out Workbench installation issues, can you also try it on the Workbench demo site at http://ryanbrainard.com/workbench? Note, that is an unsecure connection, so please use it with a test (non-production) org.

Thanks,
Ryan
sdetweilsdetweil

my org is on na11.

 

seems to do the initial connection fine.. and usually but not always gets the first notification..

then the second causes the listed failure..

 

then we're disconnected, and have to restart the workbench to recover.

1 out of 20 tries, the 1st notification causes the disconnect
altaiojokaltaiojok

Hmm...sounds like a network issue that Workbench isn't recovering from very well. I'll see what I can do to make Workbench more resilient to network blips and reconnect automatically.

 

I don't think this is the same issue, but the instance-specific issue I mentioned was fixed by replacing Line 59 in workbench/util/PhpReverseProxy.php with the following line. Can you give that a try to see if that helps at all?

$tempCookie .= trim($cookieName) . "=" . trim($cookieValue) ."; ";


Also, what browser are you using?
greentruck55greentruck55
Using firefox 3.6. Traveling til 7/28
sdetweilsdetweil

still failing with the same issue..

 

Poll Completed
successful=false
failure=true
channel=/meta/connect
request.channel=/meta/connect
request.connectionType=long-polling
request.advice.timeout=0
request.id=6
request.clientId=1r6mpu5x1d4jg1pp8j8d03rpaw
advice.reconnect=retry
advice.interval=2000
action=connect

 

host na11

 

actualy using FireFox 5.0, 64 bit

 

works once with IE9