False error message

[Sebastian42]

I have asked the following question in vain in the eMClient forum, so am now posing it here.


When I reply to an email, very often a page called OPERATIONS opens, with the following content :
<<[my email address] : Connecting to [my email address] failed [time].

[SMTP] An attempt to connect to [my email address] failed. This could be caused by temporary server unavailability or incorrect settings. Do you want to check the settings?

Open account settings Ignore >>

Opening account settings does not reveal any suggestion for improvement, so I click on IGNORE. Then the OPERATIONS window changes to say "There are no errors" and my reply gets sent and received !

Surely this is time-wasting and should be avoided. HOW can I right this ?

[JeremyNicoll]

I've never used this client, so do not know what is normal.

The website - https://www.emclient.com/ - says there are several versions (free, personal, business) and some of them include full support from the vendor.

Presumably you either never had full support, or it has expired?

What version do you have? Is it current, or old?

Have you been using this for ages and this is a sudden new problem, or has it never worked properly for you? What platform are you on?

Have you checked that your mail provider has no known problems at the moment?

Online documentation says that Settings - Logging allows you to log various things but doesn't tell me whether that includes a log showing /all/ of the to- and fro- communication between the client and your mail provider's server. I think you will need that level of logging to find out what is going wrong, eg if there is an authentication problem, or the provider's server is not responding properly and/or quickly enough.

[Sebastian42]

Thank you for your response.

I have used the free version and updated it whenever that could be done. I have had replies from eMClient previously, which shows my status is good enough to warrant a reply.

'Have you been using this for ages and this is a sudden new problem,' Yes, I have used it for years, and this problem occurs sometimes.

My platform is x32Win10ProV22H2 if I understand your question correctly.

'mail provider has no known problems at the moment' - it being Gmail.

I will try to get a log - then we can see if 'that includes a log showing /all/ of the to- and fro- communication between the client and your mail provider's server.'

I could not find a logging option.

[JeremyNicoll]

Originally Posted by Sebastian42

I could not find a logging option.

Online doc says it is in Settings - Advanced.

(Sorry for not making that clear.)

[Sebastian42]

PROGRESS - found Logging.
There were very many 'categories' for which Logging is available - I chose SMTP
The two options are to SEND or clear them.
The default destination for sending is logs@emclient.com;
However, I could also SAVE the attachment - and did so - to the desktop.
In this thread I can not find any obvious attachment option, else I would have attached the saved log.
Hope this helps.

[JeremyNicoll]

Please edit the logs first to make SURE that all occurrences of your userid & password are not visible in them. Be careful doing that, otherwise you will be telling the whole world how to access your emails.

Then if you have any cloud storage (Dropbox, OneDrive etc) upload them & post a link here.

There are also public websites where you can upload stuff, eg: "pastebin" - https://pastebin.com - I'd suggest that if you want to use that (or a similar site) you read their FAQ page first; see: https://pastebin.com/faq

[Sebastian42]

The log was saved as a ZIP. When I started to decompress it, I saw a list of dated logs from Nov 2021 till 24/07/24. The earliest are probably not relevant, but the most recent is about 8-9 months old - so of dubious value. As well there is a cef.log of half a Mb and a sqlite.log of 46Mb, both dated today.

[JeremyNicoll]

Originally Posted by Sebastian42

The log was saved as a ZIP. When I started to decompress it, I saw a list of dated logs from Nov 2021 till 24/07/24. The earliest are probably not relevant, but the most recent is about 8-9 months old - so of dubious value. As well there is a cef.log of half a Mb and a sqlite.log of 46Mb, both dated today.

If the dated logs have single dates in their names those are likely to be either when they were created or when they were last written to. You should look inside some of the apparently old ones to check they don't have recent info in them (most likely at the bottom).

Google tells me that cef.log is a standard name/format of log, but that just means the client is using logging software to create the logs & doesn't say anything useful about the contents. You should be able to look inside eg with Notepad. If you/we are unlucky it won't be readable easily - it might need software that only the vendor has.

Maybe the client used to make logs in (the old) format, but now makes the cef.log instead? Who can tell? - Not me!

Sqlite is a database format. Maybe the client keeps config data in an sqlite db; I'm not sure why you'd have an sqlite log, unless it contains either a dump/copy of all your settings - remember these logs are clearly intended to be sent to the vendor & they would need to know your - or anyone's - settings when looking into problems.

I suppose the sqlite file could also just be a set of logs; sqlite db files contain control info (for dbs) and multiple tables of data - it might just be an easy way for the client to send log text to several sub-logs at once?

You need to look inside the cef.log and sqlite.log to see, at least, if they contain meaningful-looking plain text info.

[Sebastian42]

All three files open (without 'corruption') as readable text files.

The dated log does have my gmail address a number of times, but not its password; I have not documented having an eMClient password. Even the last dated log refers only to 2024

The cef.log is numerous statements ending in "Exiting GPU process due to errors during initialization."

The sqlite.log does not have my email address, but is VERY large (46Mb) It features numerous lines like "recovered 36 frames from WAL file
C:\Users\Joe\AppData\Roaming\eM Client\Search Folders\contact_index.dat-wal" and "283 recovered 38 frames from WAL file C:\Users\Joe\AppData\Roaming\eM Client\accounts.dat-wal"

[JeremyNicoll]

WAL, or DAT-WAL is an sqlite thing for ensuring no data is lost when a program writes to a db. The entries in the sqlite.log for that possibly just mean that sqlite is working properly though they might indicate that it is having (too often) to rebuild the active db after corruption. Do you have any doubts about the disk(s) in your PC?

Leaving aside the problem you first asked about, I think you should look around the files within eMC to see if you can find sqlite logs or fragments of them. I find it hard to believe a 46 MB log file would just be from today's use, which would mean it must be partly/mostly old data ... Ask on the eMC forum if other users have lots of old log data too & if there is a proper way to get rid of the oldest stuff.


Back to your problem: the try to send, get error (timeout?), retry and it works ... could be simply a network connectivity issue. Eg, do you have a wired internet connection, or a wifi one? Wifi is often flakey, & slower, especially if you have to share it with lots of other people.

If you only send a few mails it might be that your pc's connection request doesn't find a route (from your pc to the SMTP server) in time, but when you retry the network has by then worked out how to send it, so it works.

There may be a timeout parameter within Account settings? If so, try increasing it - and again - ask on the eMC forum for advice on this.

It could be worth issuing a "traceroute" command (in a Windows cmd window) - specifying the SMTP server you're trying to use. For example, testing the route to a bbc web server:

C:\>tracert www.bbc.co.uk

Tracing route to gtm-live.pri.bbc.co.uk [212.58.237.1]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 192.168.1.1
2 6 ms 6 ms 7 ms 10.53.39.49
3 9 ms 7 ms 6 ms 80.255.198.160
4 * * * Request timed out.
5 26 ms 22 ms 19 ms tcl5-ic-4-ae5-0.network.virginmedia.net [62.252.192.246]
6 19 ms 18 ms 19 ms ae11.prc01.rbsov.bbc.co.uk [212.58.239.93]
7 19 ms 18 ms 18 ms 132.185.249.94
8 19 ms 18 ms 18 ms 212.58.237.1

Trace complete.

C:\>


This shows that traffic between me (192.168.1.1) & that web server passes (at this instant) through 6 other machines (in my ISP's network, then the BBC network) before it gets to the web server. In your case, looking at the route to the SMTP server you might find that the last line of info from tracert does not show test-data ever getting to the server.

Tracert often runs slowly. It works by asking each machine en-route to send a message back to your pc, & some systems won't do so. But it might still show a problem only when your problem occurs. The fix: speak to your ISP "about a possible routeing problem".

[Sebastian42]

Do you have any doubts about the disk(s) in your PC?
Hard Drive Sentinel reports every month that they are in top health.
As a matter of interest, surely the only disk of concern is the system disk ?

'Ask on the eMC forum'...' Will do.

'the try to send, get error (timeout?), retry and it works'
That is NOT what I wrote.
'the OPERATIONS window changes to say "There are no errors" and my reply gets sent and received !'
There is NO second attempt.

'do you have a wired internet connection' I do.
A game called Rummikub often complains about my internet, but then goes ahead and works anyway. []I think it is actually complaining about my lack of RAM - only 4Gb, because x32 can see no more].

'If you only send a few mails...' as stated above, there is NO second attempt. I suspect the traceroute advice is based on the false assumption of multiple attempts, so I wont follow it till further notice.

[JeremyNicoll]

Originally Posted by Sebastian42

Do you have any doubts about the disk(s) in your PC?
Hard Drive Sentinel reports every month that they are in top health.
As a matter of interest, surely the only disk of concern is the system disk ?

'Ask on the eMC forum'...' Will do.

'the try to send, get error (timeout?), retry and it works'
That is NOT what I wrote.
'the OPERATIONS window changes to say "There are no errors" and my reply gets sent and received !'
There is NO second attempt.

'do you have a wired internet connection' I do.
A game called Rummikub often complains about my internet, but then goes ahead and works anyway. []I think it is actually complaining about my lack of RAM - only 4Gb, because x32 can see no more].

'If you only send a few mails...' as stated above, there is NO second attempt. I suspect the traceroute advice is based on the false assumption of multiple attempts, so I wont follow it till further notice.

Disks: fine, but things like "283 recovered 38 frames from WAL file"
imply data-recovery to me and there must be a reason. You might want to read these pages:

https://stackoverflow.com/questions/...-from-wal-file

https://sqlite-users.sqlite.narkive....-journal#post2

-- note that the Richard Hipp contributing here is the designer of SQLite

-- the implication is that the SQLite db(s) are not being shut cleanly - so either
there is a disk problem or the application isn't telling SQLite to close the
db(s) when it/they are no longer in use, or you don't tell the app to close
when turning off the machine


Retry: there MUST have been one, even if you didn't explicitly ask for it. You first said:

When I reply to an email ... Connecting to [my email address] failed [time].

Then "There are no errors" ... and my reply gets sent

IT GOT SENT ... which means establishing the connection got retried, worked and then the data was sent.



As for: "I suspect the traceroute advice is based on the false assumption of multiple attempts," ... I have these comments:

1. right at the start I said "I've never used this client, so do not know what is normal."

2. what I've been doing is following a diagnostic process based on my professional IT experience.

3. neither you nor I know what the problem is, and in the absence of useful logs telling us, I am guessing & suggesting ways of getting more clues.

4. route (or general network) problems - outwith your control, probably - are a plausible cause of what you are seeing - an intermittent weird fail/retry problem. I have seen this before when a data transfer between two UK banks failed; we fixed it simply by running a "ping" just before the transfer, which forced network discovery to occur, so the route was present when needed.

[JeremyNicoll]

One other thing: there is no reason to think that there is a "false" error message.

I can understand that you find it annoying that you get told of a failure establishing a connection, but if the program didn't tell you you'd have no clues why mails did not get sent. In your case you are lucky that the retry works; suppose it didn't?

[Sebastian42]

'you are lucky that the retry works' I am, and also lucky that someone of your talent is willing to look into it.
I have not been able to do anything about my 'promises', as I was struck by migraine and have had to, recover first.

I take it you still want me to follow the traceroute advice.

'or you don't tell the app to close when turning off the machine'
That is possible - my routine is to close all programs before shutting down the PC, but then when I do shut it down, there is often a delay while it informs me that some other (mysterious thing) is still open.

It could be 'a while' before I can tend to this again.

In looking for the timeout setting you advised me to find, I came across an anomaly which I have corrected. I could not find TIMEOUT, so am stymied there, and the anomaly has to do with incoming mail, whereas my irritation is with outgoing mail.
In Synchronisation I UNTicked 'Download from POP accounts at startup', because Gmail uses IMAP. It may not produce improvement, but you never know.

I also found a statement "This application is not the default mail handler." and wonder whether I should change that. My email address is with Gmail - I don't know if Gmail is the default mail handler.

[JeremyNicoll]

Originally Posted by Sebastian42

I take it you still want me to follow the traceroute advice.

Yes, and not just once either nor just when you're using email. The only point of this is (if a route IS the problem) to find evidence of it that you can use to get your ISP to investigate.

I did that with my ISP a few years ago when tracert showed a loop; in essence traffic meant to go to some distant machine went first to (let's call it) machine B then C then G... K... L... G... K... L... G... K... L... G... .... so never got to the destination.

The fact that traffic hops from machine to machine was designed into the internet (originally ARPANET) for traffic still to get to a recipient even when intermediate systems get damaged. Every machine keeps a list of adjacent machines & of which further-off machines can be reached via them. But even in normal times the tables get out of date - eg powercuts or maintenance take machines offline, roadworks destroy cables ... Network engineers fix (broken) links, machines etc, & tweak the tables.

Whan I had my ISP fix (or ask other companies to fix) the loop, my biggest difficulty was getting to speak to someone at the ISP who knew what "network route" meant. The first-line people just went though scripts: turn your router off & on, turn your PC off & on ... etc - none of which could fix the problem.

See: https://en.wikipedia.org/wiki/ARPANET
and: https://en.wikipedia.org/wiki/Packet_switching

Originally Posted by Sebastian42

'or you don't tell the app to close when turning off the machine'
That is possible - my routine is to close all programs before shutting down the PC, but then when I do shut it down, there is often a delay while it informs me that some other (mysterious thing) is still open.

Provided you shut the mail client that should be enough.

Originally Posted by Sebastian42

I also found a statement "This application is not the default mail handler." and wonder whether I should change that. My email address is with Gmail - I don't know if Gmail is the default mail handler.

I think this is Windows telling you that eMC is not how Windows thinks it should deal with the "mailto://" protocol (used when there are clickable email-addresses on webpages & in PDFs etc), or how it would process eg ".eml" files. I do not think it is about Gmail; it's about the application ie eMC. It should not matter; you obviously do not expect eMC to start if you click on an address elsewhere - probably you just c&p the address.