Forum Replies Created
-
AuthorPosts
-
peterlaursen
ParticipantHi 😀
I saw your purchase earlier today actually! I already sent a mail to Rohit and Manoj and asked if they knew what your were doing. Your domain website is beautiful, but there is a problem with the links …
I have only a virtualized Linux here – and running SQLyog in Wine in Suse Linux in XP on a machine with 1 GB RAM is a pain, so I will let the test team check details tomorrow.
Just to the 'cut-off' thing: this may depend on how fonts in Wine are mapped to fonts in the underlying Linux system. That may differ depending on what fonts are installed. What Linux dialect is it and can you see what font is used? And do you have the MS-core fonts installed (in Linux or in Wine)?
peterlaursen
Participant@doc
I have seen you coming to the Forums form time to time. Looks like you are looking for a reply to your last post!
But I do not know what more to reply. If you think we have an issue we will need your help to reproduce it. Else we cannot debug systematically. Till now it is only more or less qualified guesswork!
peterlaursen
ParticipantSee image:
[attachment=972:slowlog28.jpg]
This is from first internal build in the 2.8 tree. This shows content that we will add in 2.8, but final design may (or rather will) become different!
The problem is that only sometimes it is possible to tell the database from the MySQL slow log. We dropped it simply because information was not always reliable! Also database display was never included in a stable verison – only betas and RC's.
But we plan to add this in a major enhancement to the 'query sniffer' (I cannot be more exact right now) planned for 3.0.
peterlaursen
ParticipantJust create a new ticket if you do not have the link to the old one!
If you have SSH option then it is also most often a better solution.
But is not the *speed* of the connection, it is the *stability*, I believe. It may be as trivial as a bad cable *anywhere* on the 'line' to the server (including the network where server is hosted) causing random disconnects. Alos it may be as issue with packet handling in some neetworking device. Just a few possibilities.
peterlaursen
Participantwriting a complete SQL parser (even if it only should support SEELCT) is a *big deal*. The server has a build-in parser. SQLyog design and concepts is basically that we do not try to 'mirror' client side what the server does.
I will not exclude the posibility of such in the future. But not for immediate future.
peterlaursen
Participantthanks for your feedback.
peterlaursen
Participantok .. there may be other errors of course, like data specific errors. I did not understand in the first place that no data were transferrred.
1) if this is a dataspecific eror we will need an example to reproduce with
2) and if it is related to the protocol/communication We will need temporary access to this server to debug this.
Can that be arranged? You can reopen your ticket (or create a new one) for privacy.
peterlaursen
ParticipantIt would be a very good idea to try from another location if you have that option. But if issue is with the server and not the 'line' it will not make difference of course!
peterlaursen
ParticipantThis is of course no TRIAL limitation. From your description the problem is simply that the HTTP connection to this server is not stable enough. Connection breaks before job has finished.
You can try another time where the server is not so busy, but it may also be an issue with the 'line' (your Telecom company, cable provider etc.). Wiht webpage display you rarely notice such issues – should a stylesheet fail to load, you would jsut refresh. SJA exits (unlike SQLyog GUI it does not try to reconnect)
But those rows that were transfered first time will not need to be transferred second time. Schedule the job to run every 10 minutes or so (it will do no harm if one job starts before the previous one has finished) and you will get to the end after some time. But best of all try direct connection or SSH-tunnel if possible.
peterlaursen
Participantthe current action you refer to is to comment like
Code:— this.. from the menu. Correctly understood?
I think adding such comment is no problem! But removing them can be tricky because finding them could be I think!
peterlaursen
Participantpeterlaursen
ParticipantI tried on this server (where webyog.com is) to experiment with permissions to the tunneller. This is a RHEL4 system.
Here
* We never thought about if before, but for years our tunneller on this server had 644 permission setting
* No write permission is required at all, but public read permission is. So also 444 works (but then you will need to change in advance should you want to delete or replace it)
* Also raising to 664 and 777 both works fine for us here. It seems that our Apache setup does not care about 'excessive privileges setting'
We will document this in a FAQ. I recall we had at least one more report with HTTP error 500 before!
If you can provide some information why a setting >644 (644) fails for you it would be nice. It could be
* An issue with a specific OS?
* Some special server directive in action?
* Some special way Apache is installed/configured in the OS?
peterlaursen
Participant“it would be a really nice improvement to show how far along the process is”
Problem is the the MySQL server returns no such information. Only when finished the server will send a status flag (successfull or not successfull). Until that the client library is waiting for the server response to the query. This is simply how MySQL client/server communication works per design.
There is no server response between start of execution and end of execution of ALTER TABLE statement.
I think we have no option to change – except that a client side time-counter is technically possible.
peterlaursen
Participant1)
The reply from your hosting support I think (after thinking about it) reflects that they did not understand it was related to tunnelling. Their reply makes sense for direction connection only I think!
2)
“it was reporting a soft exception ..” We would like to understand this. We never heard about it before! Please tell: what OS is this? It could look like some BSD security thing happening here?
3)
No write permission at all is required for the file. 444 will be enough, and I think maybe even lower will mostly work but it will depend on how Apache is configured in the OS.
peterlaursen
ParticipantI do not understand it, frankly. It must ne a very special firewall setup. Do you have phpMyAdmin of the server and dos it work? If so ask them how PMA pages can display and connect to MySQL whne the SQLyog tunneller cannot.
Please try one thing more: experiment a little with the different 'http content types' that you find in the HTTP .. advanced tab of the connection manager. (you will find a brief description in the help file).. Wiht Apache the 'standard reply is that you should always try the 'urlencode' option.
-
AuthorPosts