undefined
Skip to content
Moz logo Menu open Menu close
  • Products
    • Moz Pro
    • Moz Pro Home
    • Moz Local
    • Moz Local Home
    • STAT
    • Moz API
    • Moz API Home
    • Compare SEO Products
    • Moz Data
  • Free SEO Tools
    • Domain Analysis
    • Keyword Explorer
    • Link Explorer
    • Competitive Research
    • MozBar
    • More Free SEO Tools
  • Learn SEO
    • Beginner's Guide to SEO
    • SEO Learning Center
    • Moz Academy
    • SEO Q&A
    • Webinars, Whitepapers, & Guides
  • Blog
  • Why Moz
    • Agency Solutions
    • Enterprise Solutions
    • Small Business Solutions
    • Case Studies
    • The Moz Story
    • New Releases
  • Log in
  • Log out
  • Products
    • Moz Pro

      Your all-in-one suite of SEO essentials.

    • Moz Local

      Raise your local SEO visibility with complete local SEO management.

    • STAT

      SERP tracking and analytics for enterprise SEO experts.

    • Moz API

      Power your SEO with our index of over 44 trillion links.

    • Compare SEO Products

      See which Moz SEO solution best meets your business needs.

    • Moz Data

      Power your SEO strategy & AI models with custom data solutions.

    NEW Keyword Suggestions by Topic
    Moz Pro

    NEW Keyword Suggestions by Topic

    Learn more
  • Free SEO Tools
    • Domain Analysis

      Get top competitive SEO metrics like DA, top pages and more.

    • Keyword Explorer

      Find traffic-driving keywords with our 1.25 billion+ keyword index.

    • Link Explorer

      Explore over 40 trillion links for powerful backlink data.

    • Competitive Research

      Uncover valuable insights on your organic search competitors.

    • MozBar

      See top SEO metrics for free as you browse the web.

    • More Free SEO Tools

      Explore all the free SEO tools Moz has to offer.

    NEW Keyword Suggestions by Topic
    Moz Pro

    NEW Keyword Suggestions by Topic

    Learn more
  • Learn SEO
    • Beginner's Guide to SEO

      The #1 most popular introduction to SEO, trusted by millions.

    • SEO Learning Center

      Broaden your knowledge with SEO resources for all skill levels.

    • On-Demand Webinars

      Learn modern SEO best practices from industry experts.

    • How-To Guides

      Step-by-step guides to search success from the authority on SEO.

    • Moz Academy

      Upskill and get certified with on-demand courses & certifications.

    • SEO Q&A

      Insights & discussions from an SEO community of 500,000+.

    Unlock flexible pricing & new endpoints
    Moz API

    Unlock flexible pricing & new endpoints

    Find your plan
  • Blog
  • Why Moz
    • Small Business Solutions

      Uncover insights to make smarter marketing decisions in less time.

    • Agency Solutions

      Earn & keep valuable clients with unparalleled data & insights.

    • Enterprise Solutions

      Gain a competitive edge in the ever-changing world of search.

    • The Moz Story

      Moz was the first & remains the most trusted SEO company.

    • Case Studies

      Explore how Moz drives ROI with a proven track record of success.

    • New Releases

      Get the scoop on the latest and greatest from Moz.

    Surface actionable competitive intel
    New Feature

    Surface actionable competitive intel

    Learn More
  • Log in
    • Moz Pro
    • Moz Local
    • Moz Local Dashboard
    • Moz API
    • Moz API Dashboard
    • Moz Academy
  • Avatar
    • Moz Home
    • Notifications
    • Account & Billing
    • Manage Users
    • Community Profile
    • My Q&A
    • My Videos
    • Log Out

The Moz Q&A Forum

  • Forum
  • Questions
  • Users
  • Ask the Community

Welcome to the Q&A Forum

Browse the forum for helpful insights and fresh discussions about all things SEO.

  1. Home
  2. SEO Tactics
  3. Intermediate & Advanced SEO
  4. Recovering from robots.txt error

Moz Q&A is closed.

After more than 13 years, and tens of thousands of questions, Moz Q&A closed on 12th December 2024. Whilst we’re not completely removing the content - many posts will still be possible to view - we have locked both new posts and new replies. More details here.

Recovering from robots.txt error

Intermediate & Advanced SEO
5
11
3.2k
Loading More Posts
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as question
Log in to reply
This topic has been deleted. Only users with question management privileges can see it.
  • RikkiD22
    RikkiD22 last edited by Jun 3, 2013, 9:25 AM

    Hello,

    A client of mine is going through a bit of a crisis. A developer (at their end) added Disallow: / to the robots.txt file. Luckily the SEOMoz crawl ran a couple of days after this happened and alerted me to the error. The robots.txt file was quickly updated but the client has found the vast majority of their rankings have gone.

    It took a further 5 days for GWMT to file that the robots.txt file had been updated and since then we have "Fetched as Google" and "Submitted URL and linked pages" in GWMT.

    In GWMT it is still showing that that vast majority of pages are blocked in the "Blocked URLs" section, although the robots.txt file below it is now ok.

    I guess what I want to ask is:

    • What else is there that we can do to recover these rankings quickly?
    • What time scales can we expect for recovery?
    • More importantly has anyone had any experience with this sort of situation and is full recovery normal?

    Thanks in advance!

    1 Reply Last reply Reply Quote 0
    • BlueprintMarketing
      BlueprintMarketing @RikkiD22 last edited by Jan 5, 2014, 4:33 PM Jan 5, 2014, 4:33 PM

      Great info  Rikki

      thats goid news!

      1 Reply Last reply Reply Quote 0
      • BlueprintMarketing
        BlueprintMarketing @citricamente last edited by Jan 4, 2014, 7:08 PM Jan 4, 2014, 7:07 PM

        Hi Antonio,

        I would take a look at your entire site using

        One of my very favorite tools this tool will crawl your site and tell you if you have no follow's or other issues that would cause Google bot have trouble indexing your site.

        Simply put your sites URL in the box presented in the tool you can find in the link here

        http://www.feedthebot.com/tools/spider/

        Then use link 2

        Displays amount of links (internal, external, nofollow, image, etc.) found on webpage.

        http://www.feedthebot.com/tools/linkcount/

        You can then see if there is a no follow that might be creating a real problem inside of a page using the two URLs you should be a will to get about of this.

        Check this much of your site is you possibly can with this as it will show you A lot of information that would be very relevant as to if your site can be crawled correctly or not

        This third tool Will show you if your robots.txt file is still blocking all or part of your website the nice thing about this tool is is is built to make her about star text files however if you simply put your URL in the top and hit the upload button it will pull your robots.txt file this is very helpful when making comparisons between changes that have been made or you wish to make

        http://www.internetmarketingninjas.com/seo-tools/robots-txt-generator/

        Two check out your robot.txt file against what could be something blocking it I think that will

        http://moz.com/blog/interactive-guide-to-robots-txt

        http://moz.com/learn/seo/robotstxt

        http://tools.seobook.com/robots-txt/

        http://yoast.com/x-robots-tag-play/

        https://developers.google.com/webmasters/control-crawl-index/docs/robots_meta_tag?hl=de

        http://www.searchenginejournal.com/x-robots-tag-simple-alternate-robots-txt-meta-tag/67138/

        A citation that I hope will help you is the not too noticeable difference between allowing everything and not allowing everything simply having a / after disallow: Will tell Google that you do not want to be showing up in their search engine results

        Simply put I have the information below websites by default are set up with

        Allow: /

        Example Robots.txt Format

        Allow indexing of everything

        User-agent: *

        Disallow:

        or

        User-agent: *

        Allow: /

        Disallow indexing of everything

        User-agent: *
        Disallow: /

        Disallow indexing of a specific folder

        User-agent: *
        Disallow: /folder/

        Please remember there are multiple ways to block a website for instance

        PHP-based websites are extremely popular and if you're using a WordPress or agenda for many other

        header("X-Robots-Tag: noindex", true);

        I want to remind you what Tom Roberts  said in the first response about using Twitter I have quoted him here however you can read it at the top of the Page below the first question

        The most frequently crawled domain on the web is Twitter.  If you could legitimately get your key URLs tweeted, either by yourselves or others, this may encourage the Google crawler to revisit the URLs, and consequently re index them.  There won't be any harm SEO wise in sending tweets with your URLs, it's a quick and free method and so may be worth giving it a shot

        Hope This Helps,

        Thomas

        1 Reply Last reply Reply Quote 0
        • RikkiD22
          RikkiD22 @citricamente last edited by Jan 3, 2014, 9:54 AM Jan 3, 2014, 9:54 AM

          Hi Antonio,

          Sorry to hear you have had the same problem, due to our clients nature this error by the developer cost them a load of lost revenue.

          In answer to your questions:

          1. It took 19 days in total to recover

          2. We took everyone's advice and implemented them but I am unsure what actually helped. I think working work GWMT is the best thing for it. Make sure you submit for a re-crawl as soon as possible and see what is still blocked

          I know how scary the situation is but things will go back to normal. Its just a matter of playing the waiting game really, sorry I couldn't be of more help.

          Rikki

          BlueprintMarketing 1 Reply Last reply Jan 5, 2014, 4:33 PM Reply Quote 1
          • citricamente
            citricamente last edited by Jan 1, 2014, 9:30 PM Jan 1, 2014, 9:30 PM

            Hi Rikki,

            I know it's been some time since your post, however I just found it because a couple of weeks ago my developer did exactly the same.

            It's been 2 weeks now and our traffic is still divided by 4 compared with what it used to be. My questions are:

            1/ How long it finally took you to completely recover your previous traffic levels (if you finally did)

            2/ Did you apply any of the advices from other bloggers? What would you recommend to do from your experience?

            Thanks in advance. I am really worried at this moment, since we've got a peak campaign coming on very soon.

            Regards,

            Antonio (Citricamente)

            RikkiD22 BlueprintMarketing 2 Replies Last reply Jan 4, 2014, 7:07 PM Reply Quote 0
            • topic:timeago_earlier,7 months
            • BlueprintMarketing
              BlueprintMarketing @RikkiD22 last edited by Jun 3, 2013, 11:02 AM Jun 3, 2013, 11:02 AM

              Hi Rikki,
              I really want to say great job though with those numbers. It's always good to see somebody pulling positive ROI. Good work!

              If I may ask what type of development do specialize in if you have a specialty?

              My reason for asking is there are some excellent hosts that will allow you to run a staging server that changes everything like robots.txt back to follow and index when you hit the production button. Other hosts have similar methods.

              In fact, that might be an idea that's worth a little bit of money. A nice WordPress plug-in that gives you a constant reminder here in the development phase and does the swap then deletes itself?

              Or use a managed WordPress host if it's WordPress.
              You can do so many cool things would git these days.
              I am extremely happy you have found out there's nothing to worry about if it is simply the tags you will have your rank back before you know it.

              you can also use Webmaster tools on the manual setting and put it to Max I have done it on test sites, and the site was indexed just as well I would simply make sure I had a reminder telling me to return it to normal after.

              You should set the rel="canonical as well/

              Glad I was able to help,

              Thomas

              1 Reply Last reply Reply Quote 0
              • RikkiD22
                RikkiD22 last edited by Jun 3, 2013, 10:40 AM Jun 3, 2013, 10:40 AM

                Hi guys,

                Thanks very much for the responses. I guess my gut feeling was right that everything would come back to normal but just needed some reassurance.

                I have made real progress with this client going from an online brought in revenue of £15k per month at the start of the year to £105k last month but it is all phone based so at the moment his call centre is like a ghost town - its a shame that can happen when a developer is trying to block his own dev sub domain and ends up blocking the whole thing. Just hope it doesn't take too long.

                We will certainly try the social media route to see if that speeds things along.

                BlueprintMarketing 1 Reply Last reply Jun 3, 2013, 11:02 AM Reply Quote 0
                • BlueprintMarketing
                  BlueprintMarketing last edited by Jun 3, 2013, 10:38 AM Jun 3, 2013, 10:38 AM

                  please look and see that I updated my response I did I copied from a dictation software's writing pad and only copied a part of it when I meant to copy all of it

                  please read and let me know if I can be of help

                  sincerely,

                  Thomas

                  1 Reply Last reply Reply Quote 2
                  • BlueprintMarketing
                    BlueprintMarketing last edited by Jun 3, 2013, 10:35 AM Jun 3, 2013, 10:19 AM

                    Please forgive my 1st comment I the button too early and use the dictation software so I save it to one page then paste to another I am sincerely sorry I got this part on there without the entire thing.

                    Send me the domain either privately if you can or through this chat I would be more than happy to look into it for you. I can tell you I have made the no follow no index mistake myself showing a intern something on our own site and talk about it here below.

                    However if you are still getting problems you may want to download

                    screaming frog SEO spider

                    it only will check for 500 links for free however it gives you invaluable insight

                    It is a download and works on Mac, Windows and Linux

                    http://www.screamingfrog.co.uk/seo-spider/

                    if you want to try something web-based

                    http://www.internetmarketingninjas.com/tools/

                    http://www.internetmarketingninjas.com/broken-links-tool/

                    http://www.internetmarketingninjas.com/seo-tools/robots-txt-generator/

                    http://www.internetmarketingninjas.com/seo-tools/google-sitemap-generator/

                    I would also not hesitate to use their DNS tool to check that everything there is okay

                    Another tool or tools I would strongly recommend and you can access for free are the excellent Internet marketing ninjas

                    The words used in the metadata tags, in body text and in anchor text in external and internal links all play important roles in on page search engine optimization (SEO). The On-Page Optimization Analysis Free SEO Tool lets you quickly see the important SEO content on your webpage URL the same way a search engine spider views your data. This free SEO onpage optimization tool is multiple onpage SEO tools in one, helpful for reviewing the following onpage optimization information in the source code on the page:

                    • Metadata tool: Displays text in title tags and meta elements
                    • Keyword density tool: Reveals onpage SEO keyword statistics for linked and unlinked content
                    • Keyword optimization tool: Analyzes on page optimization by showing the number of words used in the content, including anchor text of internal and external links
                    • Link Accounting tool: Displays the number and types of links used
                    • Header check tool: Shows HTTP Status Response codes for links
                    • Source code tool: Provides quick access to on-page HTML source code

                    if you are talking about just the index and no follow

                    I can now happily say I have done this identical thing.

                    I have done the exact same thing. I can tell you I was showing somebody how to use the WordPress SEO plug-in when I got distracted and simply did not change the settings back to follow and index. So approximately 2 to 3 days later I noticed a huge loss in ranking year for the company brand name.

                    (Luckily this was mine not a clients)

                    It took approximately two days after I changed the settings back to normal follow and index them submitted my entire website to Google's Webmaster tools even clicking yes when asked the index all large change
                    before I knew it all the rankings had returned back to normal literally the keywords I was tracking returned within the normal fluctuation I see as they were in many cases sometimes better & sometimes little bit worse what I had feared they never would come back at all.

                    Sincerely,

                    Thomas

                    Believe me when I say I was extremely thankful for this and don't see why you will not get the same results with your site.

                    I hope this is a simple a mistake of just that one problem like mine that's the only thing I can give you a testimony of. I would say you have nothing to worry about. But remember to tell Google Webmaster tools I also did tell Bing but that's up to you

                    1 Reply Last reply Reply Quote 1
                    • julianhearn
                      julianhearn last edited by Jun 3, 2013, 9:37 AM Jun 3, 2013, 9:37 AM

                      Should be as quick as google re-crawls the robots.txt.

                      Best thing you can do is add a couple of links to sites that are crawled daily, to encourage google to visit your clients site as soon as possible

                      Could be:

                      • twitter
                      • facebook
                      • newspaper sites - comments
                      • and the like
                      1 Reply Last reply Reply Quote 1
                      • TomRayner
                        TomRayner last edited by Jun 3, 2013, 10:35 AM Jun 3, 2013, 9:36 AM

                        Hey there

                        I've seen this before and in almost all cases the rankings were returned to their previous state, give or take maybe 1 or 2 places (which would be normal SERP flux).

                        Unfortunately, I've found that this can often take weeks and there's no real sure-fire way of getting Google to update it quicker.  Theoretically, to speed things up you want to get the crawler revisiting the URLs more and more often.  Fresh backlinks would do this, but obviously you can't game that sort of thing for web spam reasons.  You could also try pinging devices, such as GooglePing, but I'm not convinced by their effectiveness.

                        The most frequently crawled domain on the web is Twitter.  If you could legitimately get your key URLs tweeted, either by yourselves or others, this may encourage the Google crawler to revisit the URLs, and consequently reindex them.  There won't be any harm SEO wise in sending tweets with your URLs, it's a quick and free method and so may be worth giving it a shot.

                        Hope this helps you - I've often found you can't control these things but hopefully some of these theories might work.  In the long-run, however, the rankings will return and so for normal SEO purposes, create content and links as per usual.

                        1 Reply Last reply Reply Quote 1
                        • 1 / 1
                        1 out of 11
                        • First post
                          1/11
                          Last post

                        Got a burning SEO question?

                        Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.


                        Start my free trial


                        Browse Questions

                        Explore more categories

                        • Moz Tools

                          Chat with the community about the Moz tools.

                        • SEO Tactics

                          Discuss the SEO process with fellow marketers

                        • Community

                          Discuss industry events, jobs, and news!

                        • Digital Marketing

                          Chat about tactics outside of SEO

                        • Research & Trends

                          Dive into research and trends in the search industry.

                        • Support

                          Connect on product support and feature requests.

                        • See all categories

                        Related Questions

                        • Mat_C

                          Robots.txt blocked internal resources Wordpress

                          Hi all, We've recently migrated a Wordpress website from staging to live, but the robots.txt was deleted.  I've created the following new one: User-agent: *
                          Allow: /
                          Disallow: /wp-admin/
                          Disallow: /wp-includes/
                          Disallow: /wp-content/plugins/
                          Disallow: /wp-content/cache/
                          Disallow: /wp-content/themes/
                          Allow: /wp-admin/admin-ajax.php However, in the site audit on SemRush,  I now get the mention that a lot of pages have issues with blocked internal resources in robots.txt file. These blocked internal resources are all cached and minified css elements: links, images and scripts. Does this mean that Google won't crawl some parts of these pages with blocked resources correctly and thus won't be able to follow these links and index the images? In other words, is this any cause for concern regarding SEO? Of course I can change the robots.txt again, but will urls like https://example.com/wp-content/cache/minify/df983.js end up in the index? Thanks for your thoughts!

                          Intermediate & Advanced SEO | Nov 26, 2019, 5:09 AM | Mat_C
                          2
                        • Mat_C

                          Block session id URLs with robots.txt

                          Hi, I would like to block all URLs with the parameter '?filter=' from being crawled by including them in the robots.txt. Which directive should I use: User-agent: *
                          Disallow: ?filter= or User-agent: *
                          Disallow: /?filter= In other words, is the forward slash in the beginning of the disallow directive necessary? Thanks!

                          Intermediate & Advanced SEO | Jul 9, 2019, 2:21 AM | Mat_C
                          1
                        • ThomasHarvey

                          What do you add to your robots.txt on your ecommerce sites?

                          We're looking at expanding our robots.txt, we currently don't have the ability to noindex/nofollow. We're thinking about adding the following: Checkout Basket Then possibly: Price Theme Sortby other misc filters. What do you include?

                          Intermediate & Advanced SEO | Mar 14, 2016, 4:20 PM | ThomasHarvey
                          0
                        • garrettkite

                          Images Returning 404 Error Codes. 301 Redirects?

                          We're working with a site that has gone through a lot of changes over the years - ownership, complete site redesigns, different platforms, etc. - and we are finding that there are both a lot of pages and individual images that are returning 404 error codes in the Moz crawls. We're doing 301 redirects for the pages, but what would the best course of action be for the images? The images obviously don't exist on the site anymore and are therefore returning the 404 error codes. Should we do a 301 redirect to another similar image that is on the site now or redirect the images to an actual page? Or is there another solution that I'm not considering (besides doing nothing)? We'll go through the site to make sure that there aren't any pages within the site that are still linking to those images, which is probably where the 404 errors are coming from. Based on feedback below it sounds like once we do that, leaving them alone is a good option.

                          Intermediate & Advanced SEO | Oct 8, 2014, 7:00 PM | garrettkite
                          0
                        • khi5

                          "noindex, follow" or "robots.txt" for thin content pages

                          Does anyone have any testing evidence what is better to use for pages with thin content, yet important pages to keep on a website? I am referring to content shared across multiple websites (such as e-commerce, real estate etc). Imagine a website with 300 high quality pages indexed and 5,000 thin product type pages, which are pages that would not generate relevant search traffic. Question goes: Does the interlinking value achieved by "noindex, follow" outweigh the negative of Google having to crawl all those "noindex" pages? With robots.txt one has Google's crawling focus on just the important pages that are indexed and that may give ranking a boost. Any experiments with insight to this would be great. I do get the story about "make the pages unique", "get customer reviews and comments" etc....but the above question is the important question here.

                          Intermediate & Advanced SEO | Sep 3, 2014, 6:04 PM | khi5
                          0
                        • Modi

                          Robots Disallow Backslash - Is it right command

                          Bit skeptical, as due to dynamic url and some other linkage issue, google has crawled url with backslash and asterisk character ex - www.xyz.com/\/index.php?option=com_product www.xyz.com/\"/index.php?option=com_product Now %5c is the encoded version of \ - backslash & %22 is encoded version of asterisk Need to know for command :- User-agent: *   Disallow: \As am disallowing all backslash url through this - will it only remove the backslash url which are duplicates or the entire site,

                          Intermediate & Advanced SEO | Jun 24, 2013, 11:38 PM | Modi
                          0
                        • seo123456

                          Using 2 wildcards in the robots.txt file

                          I have a URL string which I don't want to be indexed. it includes the characters _Q1 ni the middle of the string. So in the robots.txt can I use 2 wildcards in the string to take out all of the URLs with that in it?  So something like /_Q1.  Will that pickup and block every  URL with those characters in the string? Also, this is not directly of the root, but in a secondary directory, so .com/.../_Q1.  So do I have to format the robots.txt as //_Q1* as it will be in the second folder or just using /_Q1 will pickup everything no matter what folder it is on? Thanks.

                          Intermediate & Advanced SEO | Jan 20, 2012, 5:25 AM | seo123456
                          0
                        • AndrewY

                          Blocking Dynamic URLs with Robots.txt

                          Background: My e-commerce site uses a lot of layered navigation and sorting links.  While this is great for users, it ends up in a lot of URL variations of the same page being crawled by Google.  For example, a standard category page: www.mysite.com/widgets.html ...which uses a "Price" layered navigation sidebar to filter products based on price also produces the following URLs which link to the same page: http://www.mysite.com/widgets.html?price=1%2C250 http://www.mysite.com/widgets.html?price=2%2C250 http://www.mysite.com/widgets.html?price=3%2C250 As there are literally thousands of these URL variations being indexed, so I'd like to use Robots.txt to disallow these variations. Question: Is this a wise thing to do?  Or does Google take into account layered navigation links by default, and I don't need to worry. To implement, I was going to do the following in Robots.txt: User-agent: * Disallow: /*? Disallow: /*= ....which would prevent any dynamic URL with a '?" or '=' from being indexed.  Is there a better way to do this, or is this a good solution? Thank you!

                          Intermediate & Advanced SEO | May 12, 2011, 2:55 PM | AndrewY
                          1

                        Get started with Moz Pro!

                        Unlock the power of advanced SEO tools and data-driven insights.

                        Start my free trial
                        Products
                        • Moz Pro
                        • Moz Local
                        • Moz API
                        • Moz Data
                        • STAT
                        • Product Updates
                        Moz Solutions
                        • SMB Solutions
                        • Agency Solutions
                        • Enterprise Solutions
                        Free SEO Tools
                        • Domain Authority Checker
                        • Link Explorer
                        • Keyword Explorer
                        • Competitive Research
                        • Brand Authority Checker
                        • Local Citation Checker
                        • MozBar Extension
                        • MozCast
                        Resources
                        • Blog
                        • SEO Learning Center
                        • Help Hub
                        • Beginner's Guide to SEO
                        • How-to Guides
                        • Moz Academy
                        • API Docs
                        About Moz
                        • About
                        • Team
                        • Careers
                        • Contact
                        Why Moz
                        • Case Studies
                        • Testimonials
                        Get Involved
                        • Become an Affiliate
                        • MozCon
                        • Webinars
                        • Practical Marketer Series
                        • MozPod
                        Connect with us

                        Contact the Help team

                        Join our newsletter
                        Moz logo
                        © 2021 - 2025 SEOMoz, Inc., a Ziff Davis company. All rights reserved. Moz is a registered trademark of SEOMoz, Inc.
                        • Accessibility
                        • Terms of Use
                        • Privacy

                        Looks like your connection to Moz was lost, please wait while we try to reconnect.